U.S. patent application number 15/997509 was filed with the patent office on 2018-10-04 for systems and methods for replacing signal artifacts in a glucose sensor data stream.
The applicant listed for this patent is DexCom, Inc.. Invention is credited to James H. Brauker, Apurv Ullas Kamath, Ying Li, Aarthi Mahalingam.
Application Number | 20180279959 15/997509 |
Document ID | / |
Family ID | 37718472 |
Filed Date | 2018-10-04 |
United States Patent
Application |
20180279959 |
Kind Code |
A1 |
Kamath; Apurv Ullas ; et
al. |
October 4, 2018 |
SYSTEMS AND METHODS FOR REPLACING SIGNAL ARTIFACTS IN A GLUCOSE
SENSOR DATA STREAM
Abstract
Systems and methods for minimizing or eliminating transient
non-glucose related signal noise due to non-glucose rate limiting
phenomenon such as interfering species, ischemia, pH changes,
temperatures changes, known or unknown sources of mechanical,
electrical and/or biochemical noise, and the like. The system
monitors a data stream from a glucose sensor and detects signal
artifacts that have higher amplitude than electronic or
diffusion-related system noise. The system processes some or the
entire data stream continually or intermittently based at least in
part on whether the signal artifact event has occurred.
Inventors: |
Kamath; Apurv Ullas; (San
Diego, CA) ; Brauker; James H.; (Coldwater, MI)
; Mahalingam; Aarthi; (San Diego, CA) ; Li;
Ying; (San Diego, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
DexCom, Inc. |
San Diego |
CA |
US |
|
|
Family ID: |
37718472 |
Appl. No.: |
15/997509 |
Filed: |
June 4, 2018 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
15415522 |
Jan 25, 2017 |
|
|
|
15997509 |
|
|
|
|
13439675 |
Apr 4, 2012 |
9585607 |
|
|
15415522 |
|
|
|
|
11498410 |
Aug 2, 2006 |
9247901 |
|
|
13439675 |
|
|
|
|
10648849 |
Aug 22, 2003 |
8010174 |
|
|
11498410 |
|
|
|
|
11007920 |
Dec 8, 2004 |
8282549 |
|
|
10648849 |
|
|
|
|
11077739 |
Mar 10, 2005 |
8858434 |
|
|
11498410 |
|
|
|
|
60528382 |
Dec 9, 2003 |
|
|
|
60587787 |
Jul 13, 2004 |
|
|
|
60614683 |
Sep 30, 2004 |
|
|
|
60587787 |
Jul 13, 2004 |
|
|
|
60587800 |
Jul 13, 2004 |
|
|
|
60614683 |
Sep 30, 2004 |
|
|
|
60614764 |
Sep 30, 2004 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
A61B 5/7239 20130101;
A61B 5/14865 20130101; A61B 5/6848 20130101; A61B 5/14539 20130101;
A61B 5/14542 20130101; A61B 5/7203 20130101; A61B 5/742 20130101;
Y02A 90/22 20180101; Y02A 90/26 20180101; A61B 5/01 20130101; A61B
5/145 20130101; Y02A 90/10 20180101; A61B 5/7225 20130101; A61B
5/7257 20130101; A61B 2560/0252 20130101; A61B 5/726 20130101; A61B
5/14532 20130101 |
International
Class: |
A61B 5/00 20060101
A61B005/00; A61B 5/01 20060101 A61B005/01; A61B 5/145 20060101
A61B005/145; A61B 5/1486 20060101 A61B005/1486 |
Claims
1. A method of analyzing data from an analyte sensor, the method
comprising: receiving data from the analyte sensor, the data
comprising at least one sensor data point; determining whether a
signal artifact event has occurred; and processing the received
data, wherein the processing is based at least in part upon whether
the signal artifact event has occurred.
2. The method of claim 1, further comprising filtering the received
data to generate filtered data.
3. The method of claim 2, wherein determining whether a signal
artifact has occurred comprises comparing the received data with
the filtered data to obtain at least one residual.
4. The method of claim 3, wherein a signal artifact event is
determined to have occurred if the residual is exceeds a threshold
value.
5. The method of claim 4, further comprising determining whether
another signal artifact event has occurred, wherein another signal
artifact event has occurred if the residual exceeds a second
threshold value.
6. The method of claim 3, wherein a signal artifact event is
determined to have occurred if the residual is exceeds a threshold
value for a predetermined period of time or for a predetermined
amount of data.
7. The method of claim 6, wherein determining whether a signal
artifact has occurred further comprises determining whether a
predetermined number of residuals exceeds a threshold over a
predetermined period of time, or whether a predetermined amount of
data exceeds a threshold.
8. The method of claim 3, wherein determining whether a signal
artifact event has occurred further comprises determining a
differential between a first residual at a first time point and a
second residual at a second time point.
9. The method of claim 8, wherein determining whether a signal
artifact event has occurred further comprises determining whether a
predetermined number of differentials exceeds a threshold over a
predetermined period of time, or whether an amount of data exceeds
a threshold.
10. The method of claim 1, further comprising receiving reference
data from a reference analyte monitor, the reference data including
at least one reference data point.
11. The method of claim 10, wherein processing the received data
further comprises determining a reliability of the received data,
wherein processing is conducted if the signal artifact event is
determined to have not occurred.
12. The method of claim 10, further comprising matching the
reference data to substantially time corresponding received data to
form a matching data pair, wherein the reference data is matched if
the signal artifact event is determined to have not occurred.
13. The method of claim 10, further comprising including the
reference data in a calibration factor for use in calibrating the
glucose sensor, wherein the reference data is included if the
signal artifact event is determined to have not occurred.
14. The method of claim 1, further comprising prompting a user for
a reference glucose value, wherein prompting is conducted if the
signal artifact event is determined to have not occurred.
15. The method of claim 1, wherein processing the received data
comprises displaying a graphical representation of the received
data.
16. The method of claim 1, wherein processing the received data
comprises filtering the received data, wherein filtering is
conducted if the signal artifact event is determined to have
occurred.
17. The method of claim 1, further comprising filtering the
received data, wherein processing the received data comprises
displaying a graphical representation of the filtered data, wherein
processing is conducted if the signal artifact event is determined
to have occurred.
18. The method of claim 1, further comprising filtering the
received data to generate filtered data, wherein determining
whether a signal artifact event has occurred further comprises
comparing the received data with the filtered data to obtain a
residual, and wherein processing the received data comprises
utilizing the residual to modify the filtered data.
19. The method of claim 1, further comprising filtering the
received data to generate filtered data, wherein determining
whether a signal artifact event has occurred further comprises
comparing the received data with the filtered data to obtain a
residual and deriving a differential of the residual by calculating
a first derivative of the residual, and wherein processing the
received data comprises utilizing the differential to modify the
filtered data.
20. A system configured to process data from an analyte sensor, the
system comprising: a data receiving module configured to receive
sensor data from the analyte sensor, the data comprising at least
one sensor data point; a signal artifacts module configured to
detect a signal artifact in the sensor data; and a processor module
configured to process the sensor data, wherein processing is
dependent at least in part upon whether the signal artifact is
detected.
Description
INCORPORATION BY REFERENCE TO RELATED APPLICATIONS
[0001] Any and all priority claims identified in the Application
Data Sheet, or any correction thereto, are hereby incorporated by
reference under 37 CFR 1.57. This application is a continuation of
U.S. application Ser. No. 15/415,522, filed Jan. 25, 2017, which is
a continuation of U.S. application Ser. No. 13/439,675, filed Apr.
4, 2012, now U.S. Pat. No. 9,585,607, which is a continuation of
U.S. application Ser. No. 11/498,410, filed Aug. 2, 2006, now U.S.
Pat. No. 9,247,901, which is a continuation-in-part of U.S.
application Ser. No. 10/648,849, filed Aug. 22, 2003, now U.S. Pat.
No. 8,010,174. U.S. application Ser. No. 11/498,410 is a
continuation-in-part of U.S. application Ser. No. 11/007,920, filed
Dec. 8, 2004, now U.S. Pat. No. 8,282,549, which claims the benefit
of U.S. Provisional Application No. 60/528,382, filed Dec. 9, 2003,
U.S. Provisional Application No. 60/587,787, filed Jul. 13, 2004,
and U.S. Provisional Application No. 60/614,683, filed Sep. 30,
2004. U.S. application Ser. No. 11/498,410 is a
continuation-in-part of U.S. application Ser. No. 11/077,739, filed
Mar. 10, 2005, now U.S. Pat. No. 8,858,434, which claims the
benefit of U.S. Provisional Application No. 60/587,787, filed Jul.
13, 2004, U.S. Provisional Application No. 60/587,800, filed Jul.
13, 2004, U.S. Provisional Application No. 60/614,683, filed Sep.
30, 2004, and U.S. Provisional Application No. 60/614,764, filed
Sep. 30, 2004. Each of the aforementioned applications is
incorporated by reference herein in its entirety, and each is
hereby expressly made a part of this specification.
FIELD OF THE INVENTION
[0002] The present invention relates generally to systems and
methods for processing data received from a glucose sensor.
Particularly, the present invention relates to systems and methods
for detecting and processing signal artifacts, including detecting,
estimating, predicting, filtering, displaying, and otherwise
minimizing the effects of signal artifacts in a glucose sensor data
stream.
BACKGROUND OF THE INVENTION
[0003] Diabetes mellitus is a disorder in which the pancreas cannot
create sufficient insulin (Type I or insulin dependent) and/or in
which insulin is not effective (Type 2 or non-insulin dependent).
In the diabetic state, the victim suffers from high blood sugar,
which causes an array of physiological derangements (kidney
failure, skin ulcers, or bleeding into the vitreous of the eye)
associated with the deterioration of small blood vessels. A
hypoglycemic reaction (low blood sugar) is induced by an
inadvertent overdose of insulin, or after a normal dose of insulin
or glucose-lowering agent accompanied by extraordinary exercise or
insufficient food intake.
[0004] Conventionally, a diabetic person carries a self-monitoring
blood glucose (SMBG) monitor, which typically comprises
uncomfortable finger pricking methods. Due to the lack of comfort
and convenience, a diabetic will normally only measure his or her
glucose level two to four times per day. Unfortunately, these time
intervals are so far spread apart that the diabetic will likely
find out too late, sometimes incurring dangerous side effects, of a
hyperglycemic or hypoglycemic condition. In fact, it is not only
unlikely that a diabetic will take a timely SMBG value, but
additionally the diabetic will not know if their blood glucose
value is going up (higher) or down (lower) based on conventional
methods.
[0005] Consequently, a variety of transdermal and implantable
electrochemical sensors are being developed for continuous
detecting and/or quantifying blood glucose values. Many implantable
glucose sensors suffer from complications within the body and
provide only short-term and less-than-accurate sensing of blood
glucose. Similarly, transdermal sensors have run into problems in
accurately sensing and reporting back glucose values continuously
over extended periods of time. Some efforts have been made to
obtain blood glucose data from implantable devices and
retrospectively determine blood glucose trends for analysis;
however these efforts do not aid the diabetic in determining
real-time blood glucose information. Some efforts have also been
made to obtain blood glucose data from transdermal devices for
prospective data analysis, however similar problems have
occurred.
[0006] Data streams from glucose sensors are known to have some
amount of noise, caused by unwanted electronic and/or
diffusion-related system noise that degrades the quality of the
data stream. Some attempts have been made in conventional glucose
sensors to smooth the raw output data stream representative of the
concentration of blood glucose in the sample, for example by
smoothing or filtering of Gaussian, white, random, and/or other
relatively low amplitude noise in order to improve the signal to
noise ratio, and thus data output.
SUMMARY OF THE INVENTION
[0007] Systems and methods are provided that accurately detect
signal noise that is caused by substantially non-glucose reaction
rate-limiting phenomena, such as interfering species, ischemia, pH
changes, temperature changes, pressure, and stress, for example,
which are referred to herein as signal artifacts or "noise
episodes". Detecting signal artifacts and processing the sensor
data based on detection of signal artifacts provides accurate
estimated glucose measurements to a diabetic patient so that they
can proactively care for their condition to safely avoid
hyperglycemic and hypoglycemic conditions.
[0008] Accordingly, in a first aspect, a method of analyzing data
from an analyte sensor is provided, the method comprising receiving
data from the analyte sensor, the data comprising at least one
sensor data point; determining whether a signal artifact event has
occurred; and processing the received data, wherein the processing
is based at least in part upon whether the signal artifact event
has occurred.
[0009] In an embodiment of the first aspect, the method further
comprises filtering the received data to generate filtered
data.
[0010] In an embodiment of the first aspect, determining whether a
signal artifact has occurred comprises comparing the received data
with the filtered data to obtain at least one residual.
[0011] In an embodiment of the first aspect, a signal artifact
event is determined to have occurred if the residual is exceeds a
threshold value.
[0012] In an embodiment of the first aspect, the method further
comprises determining whether another signal artifact event has
occurred, wherein another signal artifact event has occurred if the
residual exceeds a second threshold value.
[0013] In an embodiment of the first aspect, a signal artifact
event is determined to have occurred if the residual is exceeds a
threshold value for a predetermined period of time or for a
predetermined amount of data.
[0014] In an embodiment of the first aspect, determining whether a
signal artifact has occurred further comprises determining whether
a predetermined number of residuals exceed a threshold over a
predetermined period of time, or whether a predetermined amount of
data exceeds a threshold.
[0015] In an embodiment of the first aspect, determining whether a
signal artifact event has occurred further comprises determining a
differential between a first residual at a first time point and a
second residual at a second time point.
[0016] In an embodiment of the first aspect, determining whether a
signal artifact event has occurred further comprises determining
whether a predetermined number of differentials exceed a threshold
over a predetermined period of time, or whether an amount of data
exceeds a threshold.
[0017] In an embodiment of the first aspect, the method further
comprises receiving reference data from a reference analyte
monitor, the reference data including at least one reference data
point.
[0018] In an embodiment of the first aspect, processing the
received data further comprises determining a reliability of the
received data, wherein processing is conducted if the signal
artifact event is determined to have not occurred.
[0019] In an embodiment of the first aspect, the method further
comprises matching the reference data to substantially time
corresponding received data to form a matching data pair, wherein
the reference data is matched if the signal artifact event is
determined to have not occurred.
[0020] In an embodiment of the first aspect, the method further
comprises including the reference data in a calibration factor for
use in calibrating the glucose sensor, wherein the reference data
is included if the signal artifact event is determined to have not
occurred.
[0021] In an embodiment of the first aspect, the method further
comprises prompting a user for a reference glucose value, wherein
prompting is conducted if the signal artifact event is determined
to have not occurred.
[0022] In an embodiment of the first aspect, processing the
received data comprises displaying a graphical representation of
the received data.
[0023] In an embodiment of the first aspect, processing the
received data comprises filtering the received data, wherein
filtering is conducted if the signal artifact event is determined
to have occurred.
[0024] In an embodiment of the first aspect, the method further
comprises filtering the received data, wherein processing the
received data comprises displaying a graphical representation of
the filtered data, wherein processing is conducted if the signal
artifact event is determined to have occurred.
[0025] In an embodiment of the first aspect, the method further
comprises filtering the received data to generate filtered data,
wherein determining whether a signal artifact event has occurred
further comprises comparing the received data with the filtered
data to obtain a residual, and wherein processing the received data
comprises utilizing the residual to modify the filtered data.
[0026] In an embodiment of the first aspect, the method further
comprises filtering the received data to generate filtered data,
wherein determining whether a signal artifact event has occurred
further comprises comparing the received data with the filtered
data to obtain a residual and deriving a differential of the
residual by calculating a first derivative of the residual, and
wherein processing the received data comprises utilizing the
differential to modify the filtered data.
[0027] In an embodiment of the first aspect, processing the
received data comprises compensating for a time lag.
[0028] In an embodiment of the first aspect, processing the
received data comprises displaying a graphical representation of
the received data.
[0029] In an embodiment of the first aspect, the received data is
an unfiltered digital signal.
[0030] In an embodiment of the first aspect, processing the
received data comprises disabling display of a graphical
representation of the received data, wherein processing is
conducted if the signal artifact event is determined to have
occurred.
[0031] In an embodiment of the first aspect, processing the
received data comprises displaying a range of glucose values,
wherein processing is conducted if the signal artifact event is
determined to have occurred.
[0032] In an embodiment of the first aspect, processing the
received data comprises displaying a graphical indication of
glucose trend, wherein processing is conducted if the signal
artifact event is determined to have occurred.
[0033] In an embodiment of the first aspect, processing the
received data comprises generating at least one estimated glucose
value and displaying a graphical representation of the estimated
glucose value, wherein processing is conducted if the signal
artifact event is determined to have occurred.
[0034] In an embodiment of the first aspect, processing the
received data comprises generating a confidence interval for at
least one estimated glucose value and displaying a graphical
representation of the confidence interval, wherein processing is
conducted if the signal artifact event is determined to have
occurred.
[0035] In a second aspect, a method for processing data from a
glucose sensor is provided, the method comprising receiving data
from the glucose sensor, the received data comprising at least one
sensor data point; displaying a graphical representation of the
data corresponding to a time period; and post-processing the
displayed graphical representation of the data corresponding to the
time period.
[0036] In an embodiment of the second aspect, post-processing is
conducted periodically.
[0037] In an embodiment of the second aspect, post-processing is
conducted substantially continuously.
[0038] In an embodiment of the second aspect, the method further
comprises determining whether a signal artifact event has occurred
and processing the received data prior to the displaying step,
wherein the processing is based at least in part upon whether the
signal artifact event has occurred.
[0039] In an embodiment of the second aspect, post-processing
comprises filtering the data to recalculate data corresponding to
the time period and displaying a graphical representation of the
recalculated data corresponding to the time period.
[0040] In an embodiment of the second aspect, the step of
post-processing comprises recalculating data corresponding to the
time period, wherein a time lag induced by real-time filtering is
substantially removed from the data corresponding to the time
period; and displaying a graphical representation of the
recalculated data corresponding to the time period.
[0041] In an embodiment of the second aspect, recalculating the
data comprises algorithmically smoothing at least one sensor data
point over a moving window, wherein the moving window comprises
time points before and after the sensor data point is obtained.
[0042] In an embodiment of the second aspect, the method further
comprises displaying a current glucose value representative of the
most recently obtained sensor data point.
[0043] In a third aspect, a system configured to process data from
an analyte sensor is provided, the system comprising a data
receiving module configured to receive sensor data from the analyte
sensor, the data comprising at least one sensor data point; a
signal artifacts module configured to detect a signal artifact in
the sensor data; and a processor module configured to process the
sensor data, wherein processing is dependent at least in part upon
whether the signal artifact is detected.
[0044] In an embodiment of the third aspect, the signal artifacts
module is configured to compare raw sensor data with filtered
sensor data to determine a residual.
[0045] In an embodiment of the third aspect, the signal artifacts
module is configured to detect a signal artifact if the residual
exceeds a threshold value.
[0046] In an embodiment of the third aspect, the signal artifacts
module is configured to detect a signal artifact if a predetermined
number of residuals exceed a threshold value for a predetermined
period of time or for a predetermined amount of data.
[0047] In an embodiment of the third aspect, the signal artifacts
module is configured to compare a first residual with a second
signal residual to determine a differential.
[0048] In an embodiment of the third aspect, the signal artifacts
module is configured to detect a signal artifact if the
differential exceeds a threshold value.
[0049] In an embodiment of the third aspect, the signal artifacts
module is configured to detect a signal artifact if a predetermined
number of differentials exceed a threshold value for a
predetermined period of time or for a predetermined amount of
data.
[0050] In an embodiment of the third aspect, the system further
comprises a reference data module configured to receive reference
data from a reference glucose monitor, the reference data
comprising at least one reference data point.
[0051] In an embodiment of the third aspect, the signal artifacts
module is configured to determine a reliability of the sensor data
if the signal artifact is detected.
[0052] In an embodiment of the third aspect, the processor module
is configured to form at least one matched data pair by matching
reference data to substantially time corresponding sensor data.
[0053] In an embodiment of the third aspect, the processor module
is configured to form a matching data pair if a signal artifact is
not detected.
[0054] In an embodiment of the third aspect, the processor module
is configured to utilize the reference data for calibrating the
glucose sensor if a signal artifact is not detected.
[0055] In an embodiment of the third aspect, the processor module
is configured to prompt a user for a reference glucose value if a
signal artifact is not detected.
[0056] In an embodiment of the third aspect, the data receiving
module is configured to receive raw sensor data.
[0057] In an embodiment of the third aspect, the raw sensor data
comprises integrated digital data.
[0058] In an embodiment of the third aspect, the processor module
is configured to display a graphical representation of the raw
sensor data if a signal artifact is not detected.
[0059] In an embodiment of the third aspect, the data receiving
module is configured to receive filtered sensor data.
[0060] In an embodiment of the third aspect, the processor module
is configured to display a graphical representation of the filtered
sensor data if a signal artifact is detected.
[0061] In an embodiment of the third aspect, the processor module
is configured to filter the sensor data.
[0062] In an embodiment of the third aspect, the processor module
is configured to display a graphical representation of the filtered
sensor data if a signal artifact is detected.
[0063] In an embodiment of the third aspect, the processor module
is configured to not display the sensor data if a signal artifact
is detected.
[0064] In an embodiment of the third aspect, the processor module
is configured to display a range of glucose values if a signal
artifact is detected.
[0065] In an embodiment of the third aspect, the processor module
is configured to display a directional indicator of glucose trend
if a signal artifact is detected.
[0066] In an embodiment of the third aspect, the processor module
is configured to display at least one estimated glucose value if a
signal artifact is detected.
[0067] In an embodiment of the third aspect, the processor module
is configured to display a confidence interval for at least one
estimated glucose value if a signal artifact is detected.
[0068] In a fourth aspect, a system configured to process data from
an analyte sensor is provided, the system comprising a data
receiving module configured to receive sensor data from the analyte
sensor, the data comprising at least one sensor data point; an
output module configured to display a substantially real-time
numerical value corresponding to a most recently received sensor
data point and a graphical representation of sensor data
corresponding to a time period; and a processor module configured
to post-process the graphical representation of the data
corresponding to the time period, wherein the output module is
configured to display the post-processed data.
[0069] In an embodiment of the fourth aspect, post-processing is
conducted periodically.
[0070] In an embodiment of the fourth aspect, post-processing is
conducted substantially continuously.
[0071] In an embodiment of the fourth aspect, the processor module
is configured to automatically post-process the graphical
representation of the data corresponding to the time period.
[0072] In an embodiment of the fourth aspect, the processor module
is configured to post-process the graphical representation of the
data corresponding to the time period responsive to a request.
[0073] In an embodiment of the fourth aspect, the output module is
configured to automatically display the post-processed graphical
representation of the data corresponding to the time period.
[0074] In an embodiment of the fourth aspect, the output module is
configured to display the post-processed graphical representation
of the data corresponding to the time period responsive to a
request.
BRIEF DESCRIPTION OF THE DRAWINGS
[0075] FIG. 1A is an exploded perspective view of a glucose sensor
in one embodiment.
[0076] FIG. 1B is side view of a distal portion of a
transcutaneously inserted sensor in one embodiment.
[0077] FIG. 2 is a block diagram that illustrates sensor
electronics in one embodiment.
[0078] FIGS. 3A to 3D are schematic views of a receiver in first,
second, third, and fourth embodiments, respectively.
[0079] FIG. 4A is a block diagram of receiver electronics in one
embodiment.
[0080] FIG. 4B is an illustration of the receiver in one embodiment
showing an analyte trend graph, including measured analyte values,
estimated analyte values, and a clinical risk zone.
[0081] FIG. 4C is an illustration of the receiver in another
embodiment showing a representation of analyte concentration and
directional trend using a gradient bar.
[0082] FIG. 4D is an illustration of the receiver in yet another
embodiment, including a screen that shows a numerical
representation of the most recent measured analyte value.
[0083] FIG. 5 is a flow chart that illustrates the process of
calibrating the sensor data in one embodiment.
[0084] FIG. 6 is a graph that illustrates a linear regression used
to calibrate the sensor data in one embodiment.
[0085] FIG. 7A is a graph that shows a raw data stream obtained
from a glucose sensor over a 4 hour time span in one example.
[0086] FIG. 7B is a graph that shows a raw data stream obtained
from a glucose sensor over a 36 hour time span in another
example.
[0087] FIG. 8 is a flow chart that illustrates the process of
detecting and replacing transient non-glucose related signal
artifacts in a data stream in one embodiment.
[0088] FIG. 9 is a graph that illustrates the correlation between
the counter electrode voltage and signal artifacts in a data stream
from a glucose sensor in one embodiment.
[0089] FIG. 10A is a circuit diagram of a potentiostat that
controls a typical three-electrode system in one embodiment.
[0090] FIG. 10B is a diagram known as Cyclic-Voltammetry (CV)
curve, which illustrates the relationship between applied potential
(V.sub.BIAS) and signal strength of the working electrode
(I.sub.SENSE) and can be used to detect signal artifacts.
[0091] FIG. 10C is a diagram showing a CV curve that illustrates an
alternative embodiment of signal artifacts detection, wherein pH
and/or temperature can be monitoring using the CV curve.
[0092] FIG. 11 is a graph and spectrogram that illustrate the
correlation between high frequency and signal artifacts observed by
monitoring the frequency content of a data stream in one
embodiment.
[0093] FIG. 12 is a graph that illustrates a data stream obtained
from a glucose sensor and a signal smoothed by trimmed linear
regression that can be used to replace some of or the entire raw
data stream in one embodiment.
[0094] FIG. 13 is a graph that illustrates a data stream obtained
from a glucose sensor and a FIR-smoothed data signal that can be
used to replace some of or the entire raw data stream in one
embodiment.
[0095] FIG. 14 is a graph that illustrates a data stream obtained
from a glucose sensor and an IR-smoothed data signal that can be
used to replace some of or the entire raw data stream in one
embodiment.
[0096] FIG. 15 is a flowchart that illustrates the process of
selectively applying signal estimation based on the severity of
signal artifacts on a data stream.
[0097] FIG. 16 is a graph that illustrates selectively applying a
signal estimation algorithm responsive to positive detection of
signal artifacts on the raw data stream.
[0098] FIG. 17 is a graph that illustrates selectively applying a
plurality of signal estimation algorithm factors responsive to a
severity of signal artifacts on the raw data stream.
[0099] FIG. 18 is a flow chart that illustrates dynamic and
intelligent estimation algorithm selection process in an
alternative embodiment.
[0100] FIG. 19 is a graph that illustrates dynamic and intelligent
estimation algorithm selection applied to a data stream in one
embodiment.
[0101] FIG. 20 is a flow chart that illustrates the process of
dynamic and intelligent estimation and evaluation of analyte values
in one embodiment.
[0102] FIG. 21 is a graph that illustrates an evaluation of the
selected estimative algorithm in one embodiment.
[0103] FIG. 22 is a flow chart that illustrates the process of
analyzing a variation of estimated future analyte value
possibilities in one embodiment.
[0104] FIG. 23 is a graph that illustrates variation analysis of
estimated glucose values in one embodiment.
[0105] FIG. 24 is a graph that illustrates variation of estimated
analyte values in another embodiment.
[0106] FIG. 25 is a flow chart that illustrates the process of
estimating, measuring, and comparing analyte values in one
embodiment.
[0107] FIG. 26 is a graph that illustrates comparison of estimated
analyte values in one embodiment.
[0108] FIG. 27 provides a flow chart that illustrates the
evaluation of reference and/or sensor data for statistical,
clinical, and/or physiological acceptability in one embodiment.
[0109] FIG. 28 is a flow chart that illustrates the evaluation of
calibrated sensor data for aberrant values in one embodiment.
[0110] FIG. 29 provides a flow chart that illustrates a
self-diagnostic of sensor data in one embodiment.
[0111] FIG. 30 is a flow chart that illustrates the process of
detecting and processing signal artifacts in certain
embodiments.
[0112] FIG. 31 is a graph that illustrates a raw data stream from a
glucose sensor for approximately 24 hours with a filtered version
of the same data stream superimposed on the same graph.
[0113] FIG. 32 is a flowchart that illustrates a method for
processing data from a glucose sensor in certain embodiments.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0114] The following description and examples illustrate some
exemplary embodiments of the disclosed invention in detail. Those
of skill in the art will recognize that there are numerous
variations and modifications of this invention that are encompassed
by its scope. Accordingly, the description of a certain exemplary
embodiment should not be deemed to limit the scope of the present
invention.
Definitions
[0115] In order to facilitate an understanding of the preferred
embodiments, a number of terms are defined below.
[0116] The term "analyte" as used herein is a broad term and is to
be given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to a
substance or chemical constituent in a biological fluid (for
example, blood, interstitial fluid, cerebral spinal fluid, lymph
fluid or urine) that can be analyzed. Analytes can include
naturally occurring substances, artificial substances, metabolites,
and/or reaction products. In some embodiments, the analyte for
measurement by the sensor heads, devices, and methods is analyte.
However, other analytes are contemplated as well, including but not
limited to acarboxyprothrombin; acylcarnitine; adenine
phosphoribosyl transferase; adenosine deaminase; albumin;
alpha-fetoprotein; amino acid profiles (arginine (Krebs cycle),
histidine/urocanic acid, homocysteine, phenylalanine/tyrosine,
tryptophan); andrenostenedione; antipyrine; arabinitol enantiomers;
arginase; benzoylecgonine (cocaine); biotinidase; biopterin;
c-reactive protein; carnitine; carnosinase; CD4; ceruloplasmin;
chenodeoxycholic acid; chloroquine; cholesterol; cholinesterase;
conjugated 1- hydroxy-cholic acid; cortisol; creatine kinase;
creatine kinase MM isoenzyme; cyclosporin A; d-penicillamine;
de-ethylchloroquine; dehydroepiandrosterone sulfate; DNA
(acetylator polymorphism, alcohol dehydrogenase, alpha
1-antitrypsin, cystic fibrosis, Duchenne/Becker muscular dystrophy,
analyte-6-phosphate dehydrogenase, hemoglobin A, hemoglobin S,
hemoglobin C, hemoglobin D, hemoglobin E, hemoglobin F, D-Punjab,
beta-thalassemia, hepatitis B virus, HCMV, HIV-1, HTLV-1, Leber
hereditary optic neuropathy, MCAD, RNA, PKU, Plasmodium vivax,
sexual differentiation, 21-deoxycortisol); desbutylhalofantrine;
dihydropteridine reductase; diptheria/tetanus antitoxin;
erythrocyte arginase; erythrocyte protoporphyrin; esterase D; fatty
acids/acylglycines; free -human chorionic gonadotropin; free
erythrocyte porphyrin; free thyroxine (FT4); free tri-iodothyronine
(FT3); fumarylacetoacetase; galactose/gal-1-phosphate;
galactose-1-phosphate uridyltransferase; gentamicin;
analyte-6-phosphate dehydrogenase; glutathione; glutathione
perioxidase; glycocholic acid; glycosylated hemoglobin;
halofantrine; hemoglobin variants; hexosaminidase A; human
erythrocyte carbonic anhydrase I; 17-alpha-hydroxyprogesterone;
hypoxanthine phosphoribosyl transferase; immunoreactive trypsin;
lactate; lead; lipoproteins ((a), B/A-1, ); lysozyme; mefloquine;
netilmicin; phenobarbitone; phenytoin; phytanic/pristanic acid;
progesterone; prolactin; prolidase; purine nucleoside
phosphorylase; quinine; reverse tri-iodothyronine (rT3); selenium;
serum pancreatic lipase; sissomicin; somatomedin C; specific
antibodies (adenovirus, anti-nuclear antibody, anti-zeta antibody,
arbovirus, Aujeszky's disease virus, dengue virus, Dracunculus
medinensis, Echinococcus granulosus, Entamoeba histolytica,
enterovirus, Giardia duodenalisa, Helicobacter pylori, hepatitis B
virus, herpes virus, HIV-1, IgE (atopic disease), influenza virus,
Leishmania donovani, leptospira, measles/mumps/rubella,
Mycobacterium leprae, Mycoplasma pneumoniae, Myoglobin, Onchocerca
volvulus, parainfluenza virus, Plasmodium falciparum, poliovirus,
Pseudomonas aeruginosa, respiratory syncytial virus, rickettsia
(scrub typhus), Schistosoma mansoni, Toxoplasma gondii, Trepenoma
pallidium, Trypanosoma cruzi/rangeli, vesicular stomatis virus,
Wuchereria bancrofti, yellow fever virus); specific antigens
(hepatitis B virus, HIV-1); succinylacetone; sulfadoxine;
theophylline; thyrotropin (TSH); thyroxine (T4); thyroxine-binding
globulin; trace elements; transferrin; UDP-galactose-4-epimerase;
urea; uroporphyrinogen I synthase; vitamin A; white blood cells;
and zinc protoporphyrin. Salts, sugar, protein, fat, vitamins, and
hormones naturally occurring in blood or interstitial fluids can
also constitute analytes in certain embodiments. The analyte can be
naturally present in the biological fluid, for example, a metabolic
product, a hormone, an antigen, an antibody, and the like.
Alternatively, the analyte can be introduced into the body, for
example, a contrast agent for imaging, a radioisotope, a chemical
agent, a fluorocarbon-based synthetic blood, or a drug or
pharmaceutical composition, including but not limited to insulin;
ethanol; cannabis (marijuana, tetrahydrocannabinol, hashish);
inhalants (nitrous oxide, amyl nitrite, butyl nitrite,
chlorohydrocarbons, hydrocarbons); cocaine (crack cocaine);
stimulants (amphetamines, methamphetamines, Ritalin, Cylert,
Preludin, Didrex, PreState, Voranil, Sandrex, Plegine); depressants
(barbituates, methaqualone, tranquilizers such as Valium, Librium,
Miltown, Serax, Equanil, Tranxene); hallucinogens (phencyclidine,
lysergic acid, mescaline, peyote, psilocybin); narcotics (heroin,
codeine, morphine, opium, meperidine, Percocet, Percodan,
Tussionex, Fentanyl, Darvon, Talwin, Lomotil); designer drugs
(analogs of fentanyl, meperidine, amphetamines, methamphetamines,
and phencyclidine, for example, Ecstasy); anabolic steroids; and
nicotine. The metabolic products of drugs and pharmaceutical
compositions are also contemplated analytes. Analytes such as
neurochemicals and other chemicals generated within the body can
also be analyzed, such as, for example, ascorbic acid, uric acid,
dopamine, noradrenaline, 3-methoxytyramine (3MT),
3,4-Dihydroxyphenylacetic acid (DOPAC), Homovanillic acid (HVA),
5-Hydroxytryptamine (5HT), and 5-Hydroxyindoleacetic acid
(FHIAA).
[0117] The term "EEPROM" as used herein is a broad term and is to
be given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
electrically erasable programmable read-only memory, which is
user-modifiable read-only memory (ROM) that can be erased and
reprogrammed (e.g., written to) repeatedly through the application
of higher than normal electrical voltage.
[0118] The term "SRAM" as used herein is a broad term and is to be
given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
static random access memory (RAM) that retains data bits in its
memory as long as power is being supplied.
[0119] The term "ROM" as used herein is a broad term and is to be
given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
read-only memory, which is a type of data storage device
manufactured with fixed contents. ROM is broad enough to include
EEPROM, for example, which is electrically erasable programmable
read-only memory (ROM).
[0120] The term "RAM" as used herein is a broad term and is to be
given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to a
data storage device for which the order of access to different
locations does not affect the speed of access. RAM is broad enough
to include SRAM, for example, which is static random access memory
that retains data bits in its memory as long as power is being
supplied.
[0121] The term "A/D Converter" as used herein is a broad term and
is to be given its ordinary and customary meaning to a person of
ordinary skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
hardware and/or software that converts analog electrical signals
into corresponding digital signals.
[0122] The terms "microprocessor" and "processor" as used herein
are broad terms and are to be given their ordinary and customary
meaning to a person of ordinary skill in the art (and are not to be
limited to a special or customized meaning), and furthermore refer
without limitation to a computer system, state machine, and the
like that performs arithmetic and logic operations using logic
circuitry that responds to and processes the basic instructions
that drive a computer.
[0123] The term "RF transceiver" as used herein is a broad term and
is to be given its ordinary and customary meaning to a person of
ordinary skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to a
radio frequency transmitter and/or receiver for transmitting and/or
receiving signals.
[0124] The term "jitter" as used herein is a broad term and is to
be given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
noise above and below the mean caused by ubiquitous noise caused by
a circuit and/or environmental effects; jitter can be seen in
amplitude, phase timing, or the width of the signal pulse.
[0125] The terms "raw data stream" and "data stream" as used herein
are broad terms and are to be given their ordinary and customary
meaning to a person of ordinary skill in the art (and are not to be
limited to a special or customized meaning), and furthermore refer
without limitation to an analog or digital signal directly related
to the measured glucose from the glucose sensor. In one example,
the raw data stream is digital data in "counts" converted by an A/D
converter from an analog signal (e.g., voltage or amps) and
includes one or more data points representative of a glucose
concentration. The terms broadly encompass a plurality of time
spaced data points from a substantially continuous glucose sensor,
which comprises individual measurements taken at time intervals
ranging from fractions of a second up to, e.g., 1, 2, or 5 minutes
or longer. In another example, the raw data stream includes an
integrated digital value, wherein the data includes one or more
data points representative of the glucose sensor signal averaged
over a time period.
[0126] The term "calibration" as used herein is a broad term and is
to be given its ordinary and customary meaning to a person of
ordinary skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
the process of determining the relationship between the sensor data
and the corresponding reference data, which can be used to convert
sensor data into meaningful values substantially equivalent to the
reference data. In some embodiments, namely, in continuous analyte
sensors, calibration can be updated or recalibrated over time as
changes in the relationship between the sensor data and reference
data occur, for example, due to changes in sensitivity, baseline,
transport, metabolism, and the like.
[0127] The terms "calibrated data" and "calibrated data stream" as
used herein are broad terms and are to be given their ordinary and
customary meaning to a person of ordinary skill in the art (and are
not to be limited to a special or customized meaning), and
furthermore refer without limitation to data that has been
transformed from its raw state to another state using a function,
for example a conversion function, to provide a meaningful value to
a user.
[0128] The terms "smoothed data" and "filtered data" as used herein
are broad terms and are to be given their ordinary and customary
meaning to a person of ordinary skill in the art (and are not to be
limited to a special or customized meaning), and furthermore refer
without limitation to data that has been modified to make it
smoother and more continuous and/or to remove or diminish outlying
points, for example, by performing a moving average of the raw data
stream. Examples of data filters include FIR (finite impulse
response), IIR (infinite impulse response), moving average filters,
and the like.
[0129] The terms "smoothing" and "filtering" as used herein are
broad terms and are to be given their ordinary and customary
meaning to a person of ordinary skill in the art (and are not to be
limited to a special or customized meaning), and furthermore refer
without limitation to modification of a set of data to make it
smoother and more continuous or to remove or diminish outlying
points, for example, by performing a moving average of the raw data
stream.
[0130] The term "algorithm" as used herein is a broad term and is
to be given its ordinary and customary meaning to a person of
ordinary skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to a
computational process (for example, programs) involved in
transforming information from one state to another, for example, by
using computer processing.
[0131] The term "matched data pairs" as used herein is a broad term
and is to be given its ordinary and customary meaning to a person
of ordinary skill in the art (and is not to be limited to a special
or customized meaning), and furthermore refers without limitation
to reference data (for example, one or more reference analyte data
points) matched with substantially time corresponding sensor data
(for example, one or more sensor data points).
[0132] The term "counts" as used herein is a broad term and is to
be given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to a
unit of measurement of a digital signal. In one example, a raw data
stream measured in counts is directly related to a voltage (e.g.,
converted by an A/D converter), which is directly related to
current from the working electrode. In another example, counter
electrode voltage measured in counts is directly related to a
voltage.
[0133] The term "sensor" as used herein is a broad term and is to
be given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
the component or region of a device by which an analyte can be
quantified.
[0134] The term "needle" as used herein is a broad term and is to
be given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to a
slender hollow instrument for introducing material into or removing
material from the body.
[0135] The terms "glucose sensor" and "member for determining the
amount of glucose in a biological sample," as used herein, are
broad terms and are used in an ordinary sense, including, without
limitation, any mechanism (e.g., enzymatic or non-enzymatic) by
which glucose can be quantified. For example, some embodiments
utilize a membrane that contains glucose oxidase that catalyzes the
conversion of oxygen and glucose to hydrogen peroxide and
gluconate, as illustrated by the following chemical reaction:
Glucose+O.sub.2.fwdarw.Gluconate+H.sub.2O.sub.2
[0136] Because for each glucose molecule metabolized, there is a
proportional change in the co-reactant O.sub.2 and the product
H.sub.2O.sub.2, one can use an electrode to monitor the current
change in either the co-reactant or the product to determine
glucose concentration.
[0137] The terms "operably connected" and "operably linked" as used
herein are broad terms and are to be given their ordinary and
customary meaning to a person of ordinary skill in the art (and are
not to be limited to a special or customized meaning), and
furthermore refer without limitation to one or more components
being linked to another component(s) in a manner that allows
transmission of signals between the components. For example, one or
more electrodes can be used to detect the amount of glucose in a
sample and convert that information into a signal, e.g., an
electrical or electromagnetic signal; the signal can then be
transmitted to an electronic circuit. In this case, the electrode
is "operably linked" to the electronic circuitry. These terms are
broad enough to include wireless connectivity.
[0138] The term "electronic circuitry" as used herein is a broad
term and is to be given its ordinary and customary meaning to a
person of ordinary skill in the art (and is not to be limited to a
special or customized meaning), and furthermore refers without
limitation to the components of a device configured to process
biological information obtained from a host. In the case of a
glucose-measuring device, the biological information is obtained by
a sensor regarding a particular glucose in a biological fluid,
thereby providing data regarding the amount of that glucose in the
fluid. U.S. Pat. Nos. 4,757,022, 5,497,772 and 4,787,398, which are
hereby incorporated by reference, describe suitable electronic
circuits that can be utilized with devices including the
biointerface membrane of a preferred embodiment.
[0139] The term "substantially" as used herein is a broad term and
is to be given its ordinary and customary meaning to a person of
ordinary skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
being largely but not necessarily wholly that which is
specified.
[0140] The term "proximal" as used herein is a broad term and is to
be given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
near to a point of reference such as an origin, a point of
attachment, or the midline of the body. For example, in some
embodiments of a glucose sensor, wherein the glucose sensor is the
point of reference, an oxygen sensor located proximal to the
glucose sensor will be in contact with or nearby the glucose sensor
such that their respective local environments are shared (e.g.,
levels of glucose, oxygen, pH, temperature, etc. are similar).
[0141] The term "distal" as used herein is a broad term and is to
be given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
spaced relatively far from a point of reference, such as an origin
or a point of attachment, or midline of the body. For example, in
some embodiments of a glucose sensor, wherein the glucose sensor is
the point of reference, an oxygen sensor located distal to the
glucose sensor will be sufficiently far from the glucose sensor
such their respective local environments are not shared (e.g.,
levels of glucose, oxygen, pH, temperature, etc. may not be
similar).
[0142] The term "domain" as used herein is a broad term and is to
be given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to a
region of the membrane system that can be a layer, a uniform or
non-uniform gradient (for example, an anisotropic region of a
membrane), or a portion of a membrane.
[0143] The terms "in vivo portion" and "distal portion" as used
herein are broad terms and are to be given their ordinary and
customary meaning to a person of ordinary skill in the art (and are
not to be limited to a special or customized meaning), and
furthermore refer without limitation to the portion of the device
(for example, a sensor) adapted for insertion into and/or existence
within a living body of a host.
[0144] The terms "ex vivo portion" and "proximal portion" as used
herein are broad terms and are to be given their ordinary and
customary meaning to a person of ordinary skill in the art (and are
not to be limited to a special or customized meaning), and
furthermore refer without limitation to the portion of the device
(for example, a sensor) adapted to remain and/or exist outside of a
living body of a host.
[0145] The term "electrochemical cell" as used herein is a broad
term and is to be given its ordinary and customary meaning to a
person of ordinary skill in the art (and is not to be limited to a
special or customized meaning), and furthermore refers without
limitation to a device in which chemical energy is converted to
electrical energy. Such a cell typically consists of two or more
electrodes held apart from each other and in contact with an
electrolyte solution. Connection of the electrodes to a source of
direct electric current renders one of them negatively charged and
the other positively charged. Positive ions in the electrolyte
migrate to the negative electrode (cathode) and there combine with
one or more electrons, losing part or all of their charge and
becoming new ions having lower charge or neutral atoms or
molecules; at the same time, negative ions migrate to the positive
electrode (anode) and transfer one or more electrons to it, also
becoming new ions or neutral particles. The overall effect of the
two processes is the transfer of electrons from the negative ions
to the positive ions, a chemical reaction.
[0146] The term "potentiostat" as used herein is a broad term and
is to be given its ordinary and customary meaning to a person of
ordinary skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
an electrical system that controls the potential between the
working and reference electrodes of a three-electrode cell at a
preset value. It forces whatever current is necessary to flow
between the working and counter electrodes to keep the desired
potential, as long as the needed cell voltage and current do not
exceed the compliance limits of the potentiostat.
[0147] The term "electrical potential" as used herein is a broad
term and is to be given its ordinary and customary meaning to a
person of ordinary skill in the art (and is not to be limited to a
special or customized meaning), and furthermore refers without
limitation to the electrical potential difference between two
points in a circuit which is the cause of the flow of a
current.
[0148] The term "host" as used herein is a broad term and is to be
given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
mammals, particularly humans.
[0149] The term "continuous analyte (or glucose) sensor" as used
herein is a broad term and is to be given its ordinary and
customary meaning to a person of ordinary skill in the art (and is
not to be limited to a special or customized meaning), and
furthermore refers without limitation to a device that continuously
or continually measures a concentration of an analyte, for example,
at time intervals ranging from fractions of a second up to, for
example, 1, 2, or 5 minutes, or longer. In one exemplary
embodiment, the continuous analyte sensor is a glucose sensor such
as described in U.S. Pat. No. 6,001,067, which is incorporated
herein by reference in its entirety.
[0150] The term "continuous analyte (or glucose) sensing" as used
herein is a broad term and is to be given its ordinary and
customary meaning to a person of ordinary skill in the art (and is
not to be limited to a special or customized meaning), and
furthermore refers without limitation to the period in which
monitoring of an analyte is continuously or continually performed,
for example, at time intervals ranging from fractions of a second
up to, for example, 1, 2, or 5 minutes, or longer.
[0151] The terms "reference analyte monitor," "reference analyte
meter," and "reference analyte sensor" as used herein are broad
terms and are to be given their ordinary and customary meaning to a
person of ordinary skill in the art (and are not to be limited to a
special or customized meaning), and furthermore refer without
limitation to a device that measures a concentration of an analyte
and can be used as a reference for the continuous analyte sensor,
for example a self-monitoring blood glucose meter (SMBG) can be
used as a reference for a continuous glucose sensor for comparison,
calibration, and the like.
[0152] The terms "sensor head" and "sensing region" as used herein
are broad terms and are to be given their ordinary and customary
meaning to a person of ordinary skill in the art (and are not to be
limited to a special or customized meaning), and furthermore refer
without limitation to the region of a monitoring device responsible
for the detection of a particular analyte. The sensing region
generally comprises a non-conductive body, a working electrode
(anode), a reference electrode (optional), and/or a counter
electrode (cathode) passing through and secured within the body
forming electrochemically reactive surfaces on the body and an
electronic connective means at another location on the body, and a
multi-domain membrane affixed to the body and covering the
electrochemically reactive surface.
[0153] The term "electrochemically reactive surface" as used herein
is a broad term and is to be given its ordinary and customary
meaning to a person of ordinary skill in the art (and is not to be
limited to a special or customized meaning), and furthermore refers
without limitation to the surface of an electrode where an
electrochemical reaction takes place. In the case of the working
electrode, the hydrogen peroxide produced by the enzyme catalyzed
reaction of the glucose being detected reacts creating a measurable
electronic current (e.g., detection of glucose utilizing glucose
oxidase produces H.sub.2O.sub.2 as a by product, H.sub.2O.sub.2
reacts with the surface of the working electrode producing two
protons (2H+), two electrons (2e.sup.-) and one molecule of oxygen
(O.sub.2) which produces the electronic current being detected). In
the case of the counter electrode, a reducible species, e.g.,
O.sub.2 is reduced at the electrode surface in order to balance the
current being generated by the working electrode.
[0154] The term "electronic connection" as used herein is a broad
term and is to be given its ordinary and customary meaning to a
person of ordinary skill in the art (and is not to be limited to a
special or customized meaning), and furthermore refers without
limitation to any electronic connection known to those in the art
that can be utilized to interface the sensor head electrodes with
the electronic circuitry of a device such as mechanical (e.g., pin
and socket) or soldered.
[0155] The term "sensing membrane" as used herein is a broad term
and is to be given its ordinary and customary meaning to a person
of ordinary skill in the art (and is not to be limited to a special
or customized meaning), and furthermore refers without limitation
to a permeable or semi-permeable membrane that can be comprised of
two or more domains and is typically constructed of materials of a
few microns thickness or more, which are permeable to oxygen and
may or may not be permeable to glucose. In one example, the sensing
membrane comprises an immobilized glucose oxidase enzyme, which
enables an electrochemical reaction to occur to measure a
concentration of glucose.
[0156] The term "biointerface membrane" as used herein is a broad
term and is to be given its ordinary and customary meaning to a
person of ordinary skill in the art (and is not to be limited to a
special or customized meaning), and furthermore refers without
limitation to a permeable membrane that can be comprised of two or
more domains and is typically constructed of materials of a few
microns thickness or more, which can be placed over the sensor body
to keep host cells (e.g., macrophages) from gaining proximity to,
and thereby damaging, the sensing membrane or forming a barrier
cell layer and interfering with the transport of glucose across the
tissue-device interface.
[0157] The term "Clarke Error Grid" as used herein is a broad term
and is to be given its ordinary and customary meaning to a person
of ordinary skill in the art (and is not to be limited to a special
or customized meaning), and furthermore refers without limitation
to an error grid analysis, which evaluates the clinical
significance of the difference between a reference glucose value
and a sensor generated glucose value, taking into account 1) the
value of the reference glucose measurement, 2) the value of the
sensor glucose measurement, 3) the relative difference between the
two values, and 4) the clinical significance of this difference.
See Clarke et al., "Evaluating Clinical Accuracy of Systems for
Self-Monitoring of Blood Glucose," Diabetes Care, Volume 10, Number
5, September-October 1987, which is incorporated by reference
herein in its entirety.
[0158] The term "physiologically feasible" as used herein is a
broad term and is to be given its ordinary and customary meaning to
a person of ordinary skill in the art (and is not to be limited to
a special or customized meaning), and furthermore refers without
limitation to the physiological parameters obtained from continuous
studies of glucose data in humans and/or animals. For example, a
maximal sustained rate of change of glucose in humans of about 4 to
5 mg/dL/min and a maximum acceleration of the rate of change of
about 0.1 to 0.2 mg/dL/min/min are deemed physiologically feasible
limits. Values outside of these limits would be considered
non-physiological and likely a result of signal error, for example.
As another example, the rate of change of glucose is lowest at the
maxima and minima of the daily glucose range, which are the areas
of greatest risk in patient treatment, thus a physiologically
feasible rate of change can be set at the maxima and minima based
on continuous studies of glucose data. As a further example, it has
been observed that the best solution for the shape of the curve at
any point along glucose signal data stream over a certain time
period (e.g., about 20 to 30 minutes) is a straight line, which can
be used to set physiological limits.
[0159] The term "ischemia" as used herein is a broad term and is to
be given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
local and temporary deficiency of blood supply due to obstruction
of circulation to a part (e.g., sensor). Ischemia can be caused by
mechanical obstruction (e.g., arterial narrowing or disruption) of
the blood supply, for example.
[0160] The term "system noise" as used herein is a broad term and
is to be given its ordinary and customary meaning to a person of
ordinary skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
unwanted electronic or diffusion-related noise which can include
Gaussian, motion-related, flicker, kinetic, or other white noise,
for example.
[0161] The terms "noise," "noise event(s)," "noise episode(s),"
"signal artifact(s)," "signal artifact event(s)," and "signal
artifact episode(s)" as used herein are broad terms and are to be
given their ordinary and customary meaning to a person of ordinary
skill in the art (and are not to be limited to a special or
customized meaning), and furthermore refer without limitation to
signal noise that is caused by substantially non-glucose related,
such as interfering species, macro- or micro-motion, ischemia, pH
changes, temperature changes, pressure, stress, or even unknown
sources of mechanical, electrical and/or biochemical noise for
example. In some embodiments, signal artifacts are transient and
characterized by a higher amplitude than system noise, and
described as "transient non-glucose related signal artifact(s) that
have a higher amplitude than system noise." In some embodiments,
noise is caused by rate-limiting (or rate-increasing) phenomena. In
some circumstances, the source of the noise is unknown.
[0162] The terms "low noise" as used herein is a broad term and is
to be given its ordinary and customary meaning to a person of
ordinary skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
noise that substantially decreases signal amplitude.
[0163] The terms "high noise" and "high spikes" as used herein are
broad terms and are to be given their ordinary and customary
meaning to a person of ordinary skill in the art (and are not to be
limited to a special or customized meaning), and furthermore refer
without limitation to noise that substantially increases signal
amplitude.
[0164] The term "frequency content" as used herein is a broad term
and is to be given its ordinary and customary meaning to a person
of ordinary skill in the art (and is not to be limited to a special
or customized meaning), and furthermore refers without limitation
to the spectral density, including the frequencies contained within
a signal and their power.
[0165] The term "spectral density" as used herein is a broad term
and is to be given its ordinary and customary meaning to a person
of ordinary skill in the art (and is not to be limited to a special
or customized meaning), and furthermore refers without limitation
to power spectral density of a given bandwidth of electromagnetic
radiation is the total power in this bandwidth divided by the
specified bandwidth. Spectral density is usually expressed in Watts
per Hertz (W/Hz).
[0166] The term "orthogonal transform" as used herein is a broad
term and is to be given its ordinary and customary meaning to a
person of ordinary skill in the art (and is not to be limited to a
special or customized meaning), and furthermore refers without
limitation to a general integral transform that is defined by
g(.alpha.)=.intg..sub.a.sup.bf(t)K(.alpha.,t)dt, where K(.alpha.,t)
represents a set of orthogonal basis functions.
[0167] The term "Fourier Transform" as used herein is a broad term
and is to be given its ordinary and customary meaning to a person
of ordinary skill in the art (and is not to be limited to a special
or customized meaning), and furthermore refers without limitation
to a technique for expressing a waveform as a weighted sum of sines
and cosines.
[0168] The term "Discrete Fourier Transform" as used herein is a
broad term and is to be given its ordinary and customary meaning to
a person of ordinary skill in the art (and is not to be limited to
a special or customized meaning), and furthermore refers without
limitation to a specialized Fourier transform where the variables
are discrete.
[0169] The term "wavelet transform" as used herein is a broad term
and is to be given its ordinary and customary meaning to a person
of ordinary skill in the art (and is not to be limited to a special
or customized meaning), and furthermore refers without limitation
to a transform which converts a signal into a series of wavelets,
which in theory allows signals processed by the wavelet transform
to be stored more efficiently than ones processed by Fourier
transform. Wavelets can also be constructed with rough edges, to
better approximate real-world signals.
[0170] The term "chronoamperometry" as used herein is a broad term
and is to be given its ordinary and customary meaning to a person
of ordinary skill in the art (and is not to be limited to a special
or customized meaning), and furthermore refers without limitation
to an electrochemical measuring technique used for electrochemical
analysis or for the determination of the kinetics and mechanism of
electrode reactions. A fast-rising potential pulse is enforced on
the working (or reference) electrode of an electrochemical cell and
the current flowing through this electrode is measured as a
function of time.
[0171] The term "pulsed amperometric detection" as used herein is a
broad term and is to be given its ordinary and customary meaning to
a person of ordinary skill in the art (and is not to be limited to
a special or customized meaning), and furthermore refers without
limitation to an electrochemical flow cell and a controller, which
applies the potentials and monitors current generated by the
electrochemical reactions. The cell can include one or multiple
working electrodes at different applied potentials. Multiple
electrodes can be arranged so that they face the chromatographic
flow independently (parallel configuration), or sequentially
(series configuration).
[0172] The term "linear regression" as used herein is a broad term
and is to be given its ordinary and customary meaning to a person
of ordinary skill in the art (and is not to be limited to a special
or customized meaning), and furthermore refers without limitation
to finding a line in which a set of data has a minimal measurement
from that line. Byproducts of this algorithm include a slope, a
y-intercept, and an R-Squared value that determine how well the
measurement data fits the line.
[0173] The term "non-linear regression" as used herein is a broad
term and is to be given its ordinary and customary meaning to a
person of ordinary skill in the art (and is not to be limited to a
special or customized meaning), and furthermore refers without
limitation to fitting a set of data to describe the relationship
between a response variable and one or more explanatory variables
in a non-linear fashion.
[0174] The term "mean" as used herein is a broad term and is to be
given its ordinary and customary meaning to a person of ordinary
skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to
the sum of the observations divided by the number of
observations.
[0175] The term "trimmed mean" as used herein is a broad term and
is to be given its ordinary and customary meaning to a person of
ordinary skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to a
mean taken after extreme values in the tails of a variable (e.g.,
highs and lows) are eliminated or reduced (e.g., "trimmed"). The
trimmed mean compensates for sensitivities to extreme values by
dropping a certain percentage of values on the tails. For example,
the 50% trimmed mean is the mean of the values between the upper
and lower quartiles. The 90% trimmed mean is the mean of the values
after truncating the lowest and highest 5% of the values. In one
example, two highest and two lowest measurements are removed from a
data set and then the remaining measurements are averaged.
[0176] The term "non-recursive filter" as used herein is a broad
term and is to be given its ordinary and customary meaning to a
person of ordinary skill in the art (and is not to be limited to a
special or customized meaning), and furthermore refers without
limitation to an equation that uses moving averages as inputs and
outputs.
[0177] The terms "recursive filter" and "auto-regressive algorithm"
as used herein are broad terms and are to be given their ordinary
and customary meaning to a person of ordinary skill in the art (and
are not to be limited to a special or customized meaning), and
furthermore refer without limitation to an equation in which
includes previous averages are part of the next filtered output.
More particularly, the generation of a series of observations
whereby the value of each observation is partly dependent on the
values of those that have immediately preceded it. One example is a
regression structure in which lagged response values assume the
role of the independent variables.
[0178] The term "signal estimation algorithm factors" as used
herein is a broad term and is to be given its ordinary and
customary meaning to a person of ordinary skill in the art (and is
not to be limited to a special or customized meaning), and
furthermore refers without limitation to one or more algorithms
that use historical and/or present signal data stream values to
estimate unknown signal data stream values. For example, signal
estimation algorithm factors can include one or more algorithms,
such as linear or non-linear regression. As another example, signal
estimation algorithm factors can include one or more sets of
coefficients that can be applied to one algorithm.
[0179] The term "variation" as used herein is a broad term and is
to be given its ordinary and customary meaning to a person of
ordinary skill in the art (and is not to be limited to a special or
customized meaning), and furthermore refers without limitation to a
divergence or amount of change from a point, line, or set of data.
In one embodiment, estimated analyte values can have a variation
including a range of values outside of the estimated analyte values
that represent a range of possibilities based on known
physiological patterns, for example.
[0180] The terms "physiological parameters" and "physiological
boundaries" as used herein are broad terms and are to be given
their ordinary and customary meaning to a person of ordinary skill
in the art (and are not to be limited to a special or customized
meaning), and furthermore refer without limitation to the
parameters obtained from continuous studies of physiological data
in humans and/or animals. For example, a maximal sustained rate of
change of glucose in humans of about 4 to 5 mg/dL/min and a maximum
acceleration of the rate of change of about 0.1 to 0.2
mg/dL/min.sup.2 are deemed physiologically feasible limits; values
outside of these limits would be considered non-physiological. As
another example, the rate of change of glucose is lowest at the
maxima and minima of the daily glucose range, which are the areas
of greatest risk in patient treatment, thus a physiologically
feasible rate of change can be set at the maxima and minima based
on continuous studies of glucose data. As a further example, it has
been observed that the best solution for the shape of the curve at
any point along glucose signal data stream over a certain time
period (for example, about 20 to 30 minutes) is a straight line,
which can be used to set physiological limits. These terms are
broad enough to include physiological parameters for any
analyte.
[0181] The term "measured analyte values" as used herein is a broad
term and is to be given its ordinary and customary meaning to a
person of ordinary skill in the art (and is not to be limited to a
special or customized meaning), and furthermore refers without
limitation to an analyte value or set of analyte values for a time
period for which analyte data has been measured by an analyte
sensor. The term is broad enough to include data from the analyte
sensor before or after data processing in the sensor and/or
receiver (for example, data smoothing, calibration, and the
like).
[0182] The term "estimated analyte values" as used herein is a
broad term and is to be given its ordinary and customary meaning to
a person of ordinary skill in the art (and is not to be limited to
a special or customized meaning), and furthermore refers without
limitation to an analyte value or set of analyte values, which have
been algorithmically extrapolated from measured analyte values.
Typically, estimated analyte values are estimated for a time period
during which no data exists. However, estimated analyte values can
also be estimated during a time period for which measured data
exists, but is to be replaced by algorithmically extrapolated (e.g.
processed or filtered) data due to noise or a time lag in the
measured data, for example.
[0183] The terms "interferants" and "interfering species" as used
herein are broad terms and are to be given their ordinary and
customary meaning to a person of ordinary skill in the art (and are
not to be limited to a special or customized meaning), and
furthermore refer without limitation to effects and/or species that
interfere with the measurement of an analyte of interest in a
sensor to produce a signal that does not accurately represent the
analyte concentration. In one example of an electrochemical sensor,
interfering species are compounds with an oxidation potential that
overlap that of the analyte to be measured, thereby producing a
false positive signal.
[0184] As employed herein, the following abbreviations apply: Eq
and Eqs (equivalents); mEq (milliequivalents); M (molar); mM
(millimolar) .mu.M (micromolar); N (Normal); mol (moles); mmol
(millimoles); .mu.mol (micromoles); nmol (nanomoles); g (grams); mg
(milligrams); .mu.g (micrograms); Kg (kilograms); L (liters); mL
(milliliters); dL (deciliters); .mu.L (microliters); cm
(centimeters); mm (millimeters); .mu.m (micrometers); nm
(nanometers); h and hr (hours); min. (minutes); s and sec.
(seconds); .degree. C. (degrees Centigrade).
[0185] Overview
[0186] The preferred embodiments relate to the use of a glucose
sensor that measures a concentration of glucose or a substance
indicative of the concentration or presence of the glucose. In some
embodiments, the glucose sensor is a continuous device, for example
a subcutaneous, transdermal, or intravascular device. In some
embodiments, the device can analyze a plurality of intermittent
blood samples. The glucose sensor can use any method of
glucose-measurement, including enzymatic, chemical, physical,
electrochemical, spectrophotometric, polarimetric, calorimetric,
iontophoretic, radiometric, and the like.
[0187] The glucose sensor can use any known method, including
invasive, minimally invasive, and non-invasive sensing techniques,
to provide a data stream indicative of the concentration of glucose
in a host. The data stream is typically a raw data signal that is
used to provide a useful value of glucose to a user, such as a
patient or doctor, who may be using the sensor. It is well known
that raw data streams typically include system noise such as
defined herein; however the preferred embodiments address the
detection and replacement of "signal artifacts" as defined herein.
Accordingly, appropriate signal estimation (e.g., filtering, data
smoothing, augmenting, projecting, and/or other methods) replace
such erroneous signals (e.g., signal artifacts) in the raw data
stream.
[0188] Glucose Sensor
[0189] The glucose sensor can be any device capable of measuring
the concentration of glucose. One exemplary embodiment is described
below, which utilizes an implantable glucose sensor. However, it
should be understood that the devices and methods described herein
can be applied to any device capable of detecting a concentration
of glucose and providing an output signal that represents the
concentration of glucose.
[0190] In one preferred embodiment, the analyte sensor is an
implantable glucose sensor, such as described with reference to
U.S. Pat. No. 6,001,067 and U.S. Publication No.
US-2005-0027463-A1. In another preferred embodiment, the analyte
sensor is a transcutaneous glucose sensor, such as described with
reference to U.S. Publication No. US-2006-0020187-A1. In one
alternative embodiment, the continuous glucose sensor comprises a
transcutaneous sensor such as described in U.S. Pat. No. 6,565,509
to Say et al., for example. In another alternative embodiment, the
continuous glucose sensor comprises a subcutaneous sensor such as
described with reference to U.S. Pat. No. 6,579,690 to Bonnecaze et
al. or U.S. Pat. No. 6,484,046 to Say et al., for example. In
another alternative embodiment, the continuous glucose sensor
comprises a refillable subcutaneous sensor such as described with
reference to U.S. Pat. No. 6,512,939 to Colvin et al., for example.
In another alternative embodiment, the continuous glucose sensor
comprises an intravascular sensor such as described with reference
to U.S. Pat. No. 6,477,395 to Schulman et al., for example. In
another alternative embodiment, the continuous glucose sensor
comprises an intravascular sensor such as described with reference
to U.S. Pat. No. 6,424,847 to Mastrototaro et al.
[0191] FIG. 1A is an exploded perspective view of one exemplary
embodiment comprising an implantable glucose sensor 10 that
utilizes amperometric electrochemical sensor technology to measure
glucose concentration. In this exemplary embodiment, a body 12 and
head 14 house the electrodes 16 and sensor electronics, which are
described in more detail below with reference to FIG. 2. Three
electrodes 16 are operably connected to the sensor electronics
(FIG. 2) and are covered by a sensing membrane 17 and a
biointerface membrane 18, which are attached by a clip 19.
[0192] In one embodiment, the three electrodes 16, which protrude
through the head 14, include a platinum working electrode, a
platinum counter electrode, and a silver/silver chloride reference
electrode. The top ends of the electrodes are in contact with an
electrolyte phase (not shown), which is a free-flowing fluid phase
disposed between the sensing membrane 17 and the electrodes 16. The
sensing membrane 17 includes an enzyme, e.g., glucose oxidase,
which covers the electrolyte phase. The biointerface membrane 18
covers the sensing membrane 17 and serves, at least in part, to
protect the sensor 10 from external forces that can result in
environmental stress cracking of the sensing membrane 17.
[0193] In the illustrated embodiment, the counter electrode is
provided to balance the current generated by the species being
measured at the working electrode. In the case of a glucose oxidase
based glucose sensor, the species being measured at the working
electrode is H.sub.2O.sub.2. Glucose oxidase catalyzes the
conversion of oxygen and glucose to hydrogen peroxide and gluconate
according to the following reaction:
Glucose+O.sub.2.fwdarw.Gluconate+H.sub.2O.sub.2
[0194] The change in H.sub.2O.sub.2 can be monitored to determine
glucose concentration because for each glucose molecule
metabolized, there is a proportional change in the product
H.sub.2O.sub.2. Oxidation of H.sub.2O.sub.2 by the working
electrode is balanced by reduction of ambient oxygen, enzyme
generated H.sub.2O.sub.2, or other reducible species at the counter
electrode. The H.sub.2O.sub.2 produced from the glucose oxidase
reaction further reacts at the surface of working electrode and
produces two protons (2H+), two electrons (2e.sup.-), and one
oxygen molecule (O.sub.2).
[0195] FIG. 1B is side view of a distal portion of a
transcutaneously-inserted sensor 100 in one embodiment, showing
working and reference electrodes. In preferred embodiments, the
sensor 100 is formed from a working electrode 244 and a reference
electrode 246 helically wound around the working electrode 244. An
insulator 245 is disposed between the working and reference
electrodes to provide necessary electrical insulation therebetween.
Certain portions of the electrodes are exposed to enable
electrochemical reaction thereon, for example, a window 243 can be
formed in the insulator to expose a portion of the working
electrode 244 for electrochemical reaction.
[0196] In preferred embodiments, each electrode is formed from a
fine wire with a diameter of from about 0.001 or less to about
0.010 inches or more, for example, and is formed from, e.g., a
plated insulator, a plated wire, or bulk electrically conductive
material. Although the illustrated electrode configuration and
associated text describe one preferred method of forming a
transcutaneous sensor, a variety of known transcutaneous sensor
configurations can be employed with the transcutaneous analyte
sensor system of the preferred embodiments, such as are described
in U.S. Pat. No. 6,695,860 to Ward et al., U.S. Pat. No. 6,565,509
to Say et al., U.S. Pat. No. 6,248,067 to Causey III, et al., and
U.S. Pat. No. 6,514,718 to Heller et al.
[0197] In preferred embodiments, the working electrode comprises a
wire formed from a conductive material, such as platinum,
platinum-iridium, palladium, graphite, gold, carbon, conductive
polymer, alloys, and the like. Although the electrodes can by
formed by a variety of manufacturing techniques (bulk metal
processing, deposition of metal onto a substrate, and the like), it
can be advantageous to form the electrodes from plated wire (e.g.,
platinum on steel wire) or bulk metal (e.g., platinum wire). It is
believed that electrodes formed from bulk metal wire provide
superior performance (e.g., in contrast to deposited electrodes),
including increased stability of assay, simplified
manufacturability, resistance to contamination (e.g., which can be
introduced in deposition processes), and improved surface reaction
(e.g., due to purity of material) without peeling or
delamination.
[0198] The working electrode 244 is configured to measure the
concentration of an analyte. In an enzymatic electrochemical sensor
for detecting glucose, for example, the working electrode measures
the hydrogen peroxide produced by an enzyme catalyzed reaction of
the analyte being detected and creates a measurable electronic
current. For example, in the detection of glucose wherein glucose
oxidase produces hydrogen peroxide as a byproduct, hydrogen
peroxide reacts with the surface of the working electrode producing
two protons (2H.sup.+), two electrons (2e.sup.-) and one molecule
of oxygen (O.sub.2), which produces the electronic current being
detected.
[0199] In preferred embodiments, the working electrode 244 is
covered with an insulating material 45, for example, a
non-conductive polymer. Dip-coating, spray-coating,
vapor-deposition, or other coating or deposition techniques can be
used to deposit the insulating material on the working electrode.
In one embodiment, the insulating material comprises parylene,
which can be an advantageous polymer coating for its strength,
lubricity, and electrical insulation properties. Generally,
parylene is produced by vapor deposition and polymerization of
para-xylylene (or its substituted derivatives). However, any
suitable insulating material can be used, for example, fluorinated
polymers, polyethyleneterephthalate, polyurethane, polyimide, other
nonconducting polymers, and the like. Glass or ceramic materials
can also be employed. Other materials suitable for use include
surface energy modified coating systems such as are marketed under
the trade names AMC18, AMC148, AMC141, and AMC321 by Advanced
Materials Components Express of Bellafonte, Pa. In some alternative
embodiments, however, the working electrode may not require a
coating of insulator.
[0200] The reference electrode 246, which can function as a
reference electrode alone, or as a dual reference and counter
electrode, is formed from silver, silver/silver chloride, and the
like. Preferably, the reference electrode 246 is juxtapositioned
and/or twisted with or around the working electrode 244; however
other configurations are also possible. In the illustrated
embodiments, the reference electrode 246 is helically wound around
the working electrode 244. The assembly of wires is then optionally
coated or adhered together with an insulating material, similar to
that described above, so as to provide an insulating
attachment.
[0201] In embodiments wherein an outer insulator is disposed, a
portion of the coated assembly structure can be stripped or
otherwise removed, for example, by hand, excimer lasing, chemical
etching, laser ablation, grit-blasting (e.g., with sodium
bicarbonate or other suitable grit), and the like, to expose the
electroactive surfaces. Alternatively, a portion of the electrode
can be masked prior to depositing the insulator in order to
maintain an exposed electroactive surface area. In one exemplary
embodiment, grit blasting is implemented to expose the
electroactive surfaces, preferably utilizing a grit material that
is sufficiently hard to ablate the polymer material, while being
sufficiently soft so as to minimize or avoid damage to the
underlying metal electrode (e.g., a platinum electrode). Although a
variety of "grit" materials can be used (e.g., sand, talc, walnut
shell, ground plastic, sea salt, and the like), in some preferred
embodiments, sodium bicarbonate is an advantageous grit-material
because it is sufficiently hard to ablate, e.g., a parylene coating
without damaging, e.g., an underlying platinum conductor. One
additional advantage of sodium bicarbonate blasting includes its
polishing action on the metal as it strips the polymer layer,
thereby eliminating a cleaning step that might otherwise be
necessary.
[0202] In the embodiment illustrated in FIG. 1B, a radial window
243 is formed through the insulating material 245 to expose a
circumferential electroactive surface of the working electrode.
Additionally, sections 241 of electroactive surface of the
reference electrode are exposed. For example, the 241 sections of
electroactive surface can be masked during deposition of an outer
insulating layer or etched after deposition of an outer insulating
layer.
[0203] In some applications, cellular attack or migration of cells
to the sensor can cause reduced sensitivity and/or function of the
device, particularly after the first day of implantation. However,
when the exposed electroactive surface is distributed
circumferentially about the sensor (e.g., as in a radial window),
the available surface area for reaction can be sufficiently
distributed so as to minimize the effect of local cellular invasion
of the sensor on the sensor signal. Alternatively, a tangential
exposed electroactive window can be formed, for example, by
stripping only one side of the coated assembly structure. In other
alternative embodiments, the window can be provided at the tip of
the coated assembly structure such that the electroactive surfaces
are exposed at the tip of the sensor. Other methods and
configurations for exposing electroactive surfaces can also be
employed.
[0204] In some embodiments, the working electrode has a diameter of
from about 0.001 inches or less to about 0.010 inches or more,
preferably from about 0.002 inches to about 0.008 inches, and more
preferably from about 0.004 inches to about 0.005 inches. The
length of the window can be from about 0.1 mm (about 0.004 inches)
or less to about 2 mm (about 0.078 inches) or more, and preferably
from about 0.5 mm (about 0.02 inches) to about 0.75 mm (0.03
inches). In such embodiments, the exposed surface area of the
working electrode is preferably from about 0.000013 in.sup.2
(0.0000839 cm.sup.2) or less to about 0.0025 in.sup.2 (0.016129
cm.sup.2) or more (assuming a diameter of from about 0.001 inches
to about 0.010 inches and a length of from about 0.004 inches to
about 0.078 inches). The preferred exposed surface area of the
working electrode is selected to produce an analyte signal with a
current in the picoAmp range, such as is described in more detail
elsewhere herein. However, a current in the picoAmp range can be
dependent upon a variety of factors, for example the electronic
circuitry design (e.g., sample rate, current draw, A/D converter
bit resolution, etc.), the membrane system (e.g., permeability of
the analyte through the membrane system), and the exposed surface
area of the working electrode. Accordingly, the exposed
electroactive working electrode surface area can be selected to
have a value greater than or less than the above-described ranges
taking into consideration alterations in the membrane system and/or
electronic circuitry. In preferred embodiments of a glucose sensor,
it can be advantageous to minimize the surface area of the working
electrode while maximizing the diffusivity of glucose in order to
optimize the signal-to-noise ratio while maintaining sensor
performance in both high and low glucose concentration ranges.
[0205] In some alternative embodiments, the exposed surface area of
the working (and/or other) electrode can be increased by altering
the cross-section of the electrode itself. For example, in some
embodiments the cross-section of the working electrode can be
defined by a cross, star, cloverleaf, ribbed, dimpled, ridged,
irregular, or other non-circular configuration; thus, for any
predetermined length of electrode, a specific increased surface
area can be achieved (as compared to the area achieved by a
circular cross-section). Increasing the surface area of the working
electrode can be advantageous in providing an increased signal
responsive to the analyte concentration, which in turn can be
helpful in improving the signal-to-noise ratio, for example.
[0206] In some alternative embodiments, additional electrodes can
be included within the assembly, for example, a three-electrode
system (working, reference, and counter electrodes) and/or an
additional working electrode (e.g., an electrode which can be used
to generate oxygen, which is configured as a baseline subtracting
electrode, or which is configured for measuring additional
analytes). U.S. Publication No. US-2005-0161346-A1 and U.S.
Publication No. US-2005-0143635-A1 describe some systems and
methods for implementing and using additional working, counter,
and/or reference electrodes. In one implementation wherein the
sensor comprises two working electrodes, the two working electrodes
are juxtapositioned (e.g., extend parallel to each other), around
which the reference electrode is disposed (e.g., helically wound).
In some embodiments wherein two or more working electrodes are
provided, the working electrodes can be formed in a double-,
triple-, quad-, etc. helix configuration along the length of the
sensor (for example, surrounding a reference electrode, insulated
rod, or other support structure). The resulting electrode system
can be configured with an appropriate membrane system, wherein the
first working electrode is configured to measure a first signal
comprising glucose and baseline and the additional working
electrode is configured to measure a baseline signal consisting of
baseline only (e.g., configured to be substantially similar to the
first working electrode without an enzyme disposed thereon). In
this way, the baseline signal can be subtracted from the first
signal to produce a glucose-only signal that is substantially not
subject to fluctuations in the baseline and/or interfering species
on the signal.
[0207] Although the preferred embodiments illustrate one electrode
configuration including one bulk metal wire helically wound around
another bulk metal wire, other electrode configurations are also
contemplated. In an alternative embodiment, the working electrode
comprises a tube with a reference electrode disposed or coiled
inside, including an insulator therebetween. Alternatively, the
reference electrode comprises a tube with a working electrode
disposed or coiled inside, including an insulator therebetween. In
another alternative embodiment, a polymer (e.g., insulating) rod is
provided, wherein the electrodes are deposited (e.g.,
electro-plated) thereon. In yet another alternative embodiment, a
metallic (e.g., steel) rod is provided, coated with an insulating
material, onto which the working and reference electrodes are
deposited. In yet another alternative embodiment, one or more
working electrodes are helically wound around a reference
electrode.
[0208] Preferably, the electrodes and membrane systems of the
preferred embodiments are coaxially formed, namely, the electrodes
and/or membrane system all share the same central axis. While not
wishing to be bound by theory, it is believed that a coaxial design
of the sensor enables a symmetrical design without a preferred bend
radius. Namely, in contrast to prior art sensors comprising a
substantially planar configuration that can suffer from regular
bending about the plane of the sensor, the coaxial design of the
preferred embodiments do not have a preferred bend radius and
therefore are not subject to regular bending about a particular
plane (which can cause fatigue failures and the like). However,
non-coaxial sensors can be implemented with the sensor system of
the preferred embodiments.
[0209] In addition to the above-described advantages, the coaxial
sensor design of the preferred embodiments enables the diameter of
the connecting end of the sensor (proximal portion) to be
substantially the same as that of the sensing end (distal portion)
such that the needle is able to insert the sensor into the host and
subsequently slide back over the sensor and release the sensor from
the needle, without slots or other complex multi-component
designs.
[0210] In one such alternative embodiment, the two wires of the
sensor are held apart and configured for insertion into the host in
proximal but separate locations. The separation of the working and
reference electrodes in such an embodiment can provide additional
electrochemical stability with simplified manufacture and
electrical connectivity. It is appreciated by one skilled in the
art that a variety of electrode configurations can be implemented
with the preferred embodiments.
[0211] Preferably, a membrane system is deposited over the
electroactive surfaces of the sensor 100 and includes a plurality
of domains or layers. The membrane system may be deposited on the
exposed electroactive surfaces using known thin film techniques
(for example, spraying, electro-depositing, dipping, and the like).
In one exemplary embodiment, each domain is deposited by dipping
the sensor into a solution and drawing out the sensor at a speed
that provides the appropriate domain thickness. In general, the
membrane system may be disposed over (deposited on) the
electroactive surfaces using methods appreciated by one skilled in
the art.
[0212] In one exemplary embodiment, the sensor is an enzyme-based
electrochemical sensor, wherein the glucose-measuring working
electrode measures the hydrogen peroxide produced by the enzyme
catalyzed reaction of glucose being detected and creates a
measurable electronic current (for example, detection of glucose
utilizing glucose oxidase produces H.sub.2O.sub.2 peroxide as a by
product, H.sub.2O.sub.2 reacts with the surface of the working
electrode producing two protons (2H.sup.+), two electrons
(2e.sup.-) and one molecule of oxygen (02) which produces the
electronic current being detected), such as described in more
detail above and as is appreciated by one skilled in the art.
Typically, the working and reference electrodes operatively connect
with sensor electronics, such as described in more detail elsewhere
herein. Additional aspects of the above-described transcutaneously
inserted sensor can be found in co-pending U.S. Publication No.
US-2006-0020187-A1.
[0213] In some embodiments (e.g., sensors such as illustrated in
FIGS. 1A and 1B), a potentiostat is employed to monitor the
electrochemical reaction at the electrochemical cell. The
potentiostat applies a constant potential to the working and
reference electrodes to determine a current value. The current that
is produced at the working electrode (and flows through the
circuitry to the counter electrode) is proportional to the amount
of H.sub.2O.sub.2 that diffuses to the working electrode.
Accordingly, a raw signal can be produced that is representative of
the concentration of glucose in the user's body, and therefore can
be utilized to estimate a meaningful glucose value, such as
described herein.
[0214] One problem with raw data stream output of enzymatic glucose
sensors such as described above is caused by transient non-glucose
reaction rate-limiting phenomenon. For example, if oxygen is
deficient, relative to the amount of glucose, then the enzymatic
reaction will be limited by oxygen rather than glucose.
Consequently, the output signal will be indicative of the oxygen
concentration rather than the glucose concentration, producing
erroneous signals. Other non-glucose reaction rate-limiting
phenomenon could include interfering species, temperature and/or pH
changes, or even unknown sources of mechanical, electrical and/or
biochemical noise, for example. Accordingly, reduction of signal
noise, and particularly replacement of transient non-glucose
related signal artifacts in the data stream that have a higher
amplitude than system noise, can be performed in the sensor and/or
in the receiver, such as described in more detail below in the
sections entitled "Signal Artifacts Detection" and "Signal
Artifacts Replacement," for example.
[0215] FIG. 2 is a block diagram that illustrates one possible
configuration of the sensor electronics in one embodiment. In this
embodiment, a potentiostat 20 is shown, which is operatively
connected to an electrode system (FIG. 1A or 1B) and provides a
voltage to the electrodes, which biases the sensor to enable
measurement of a current value indicative of the analyte
concentration in the host (also referred to as the analog portion).
In some embodiments, the potentiostat includes a resistor (not
shown) that translates the current into voltage. In some
alternative embodiments, a current to frequency converter is
provided that is configured to continuously integrate the measured
current, for example, using a charge counting device. In the
illustrated embodiment, an A/D converter 21 digitizes the analog
signal into "counts" for processing. Accordingly, the resulting raw
data stream in counts is directly related to the current measured
by the potentiostat 20.
[0216] A processor module 22 is the central control unit that
controls the processing of the sensor electronics. In some
embodiments, the processor module includes a microprocessor,
however a computer system other than a microprocessor can be used
to process data as described herein, for example an ASIC can be
used for some or all of the sensor's central processing. The
processor typically provides semi-permanent storage of data, for
example, storing data such as sensor identifier (ID) and
programming to process data streams (for example, programming for
data smoothing and/or replacement of signal artifacts such as is
described in more detail elsewhere herein). The processor
additionally can be used for the system's cache memory, for example
for temporarily storing recent sensor data. In some embodiments,
the processor module comprises memory storage components such as
ROM, RAM, dynamic-RAM, static-RAM, non-static RAM, EEPROM,
rewritable ROMs, flash memory, and the like. In one exemplary
embodiment, EEPROM 23 provides semi-permanent storage of data, for
example, storing data such as sensor identifier (ID) and
programming to process data streams (e.g., programming for signal
artifacts detection and/or replacement such as described elsewhere
herein). In one exemplary embodiment, SRAM 24 can be used for the
system's cache memory, for example for temporarily storing recent
sensor data.
[0217] In some embodiments, the processor module comprises a
digital filter, for example, an IIR or FIR filter, configured to
smooth the raw data stream from the A/D converter. Generally,
digital filters are programmed to filter data sampled at a
predetermined time interval (also referred to as a sample rate). In
some embodiments, wherein the potentiostat is configured to measure
the analyte at discrete time intervals, these time intervals
determine the sample rate of the digital filter. In some
alternative embodiments, wherein the potentiostat is configured to
continuously measure the analyte, for example, using a
current-to-frequency converter, the processor module can be
programmed to request a digital value from the A/D converter at a
predetermined time interval, also referred to as the acquisition
time. In these alternative embodiments, the values obtained by the
processor are advantageously averaged over the acquisition time due
the continuity of the current measurement. Accordingly, the
acquisition time determines the sample rate of the digital filter.
In preferred embodiments, the processor module is configured with a
programmable acquisition time, namely, the predetermined time
interval for requesting the digital value from the A/D converter is
programmable by a user within the digital circuitry of the
processor module. An acquisition time of from about 2 seconds to
about 512 seconds is preferred; however any acquisition time can be
programmed into the processor module. A programmable acquisition
time is advantageous in optimizing noise filtration, time lag, and
processing/battery power.
[0218] Preferably, the processor module is configured to build the
data packet for transmission to an outside source, for example, an
RF transmission to a receiver as described in more detail below.
Generally, the data packet comprises a plurality of bits that can
include a sensor ID code, raw data, filtered data, and/or error
detection or correction. The processor module can be configured to
transmit any combination of raw and/or filtered data.
[0219] A battery 25 is operatively connected to the processor 22
and provides the necessary power for the sensor (e.g., 10 or 100).
In one embodiment, the battery is a Lithium Manganese Dioxide
battery, however any appropriately sized and powered battery can be
used (e.g., AAA, Nickel-cadmium, Zinc-carbon, Alkaline, Lithium,
Nickel-metal hydride, Lithium-ion, Zinc-air, Zinc-mercury oxide,
Silver-zinc, or hermetically-sealed). In some embodiments the
battery is rechargeable. In some embodiments, a plurality of
batteries can be used to power the system. In yet other
embodiments, the receiver can be transcutaneously powered via an
inductive coupling, for example. A Quartz Crystal 26 is operatively
connected to the processor 22 and maintains system time for the
computer system as a whole.
[0220] An RF module, (e.g., an RF Transceiver) 27 is operably
connected to the processor 22 and transmits the sensor data from
the sensor (e.g., 10 or 100) to a receiver (see FIGS. 3 and 4).
Although an RF transceiver is shown here, some other embodiments
can include a wired rather than wireless connection to the
receiver. A second quartz crystal 28 provides the system time for
synchronizing the data transmissions from the RF transceiver. It is
noted that the transceiver 27 can be substituted with a transmitter
in other embodiments. In some alternative embodiments, however,
other mechanisms, such as optical, infrared radiation (IR),
ultrasonic, and the like, can be used to transmit and/or receive
data.
[0221] In some embodiments, a Signal Artifacts Detector 29 is
provided that includes one or more of the following: an oxygen
detector 29a, a pH detector 29b, a temperature detector 29c, and a
pressure/stress detector 29d, which is described in more detail
with reference to signal artifacts detection. It is noted that in
some embodiments the signal artifacts detector 29 is a separate
entity (e.g., temperature detector) operatively connected to the
processor, while in other embodiments, the signal artifacts
detector is a part of the processor and utilizes readings from the
electrodes, for example, to detect ischemia and other signal
artifacts. Although the above description is focused on an
embodiment of the Signal Artifacts Detector within the sensor, some
embodiments provide for systems and methods for detecting signal
artifacts in the sensor and/or receiver electronics (e.g.,
processor module) as described in more detail elsewhere herein.
[0222] Receiver
[0223] FIGS. 3A to 3D are schematic views of a receiver 30
including representations of estimated glucose values on its user
interface in first, second, third, and fourth embodiments,
respectively. The receiver 30 comprises systems to receive,
process, and display sensor data from the glucose sensor (e.g., 10
or 100), such as described herein. Particularly, the receiver 30
can be a pager-sized device, for example, and comprise a user
interface that has a plurality of buttons 32 and a liquid crystal
display (LCD) screen 34, and which can optionally include a
backlight. In some embodiments, the user interface can also include
a keyboard, a speaker, and a vibrator, as described below with
reference to FIG. 4A.
[0224] FIG. 3A illustrates a first embodiment wherein the receiver
30 shows a numeric representation of the estimated glucose value on
its user interface, which is described in more detail elsewhere
herein.
[0225] FIG. 3B illustrates a second embodiment wherein the receiver
30 shows an estimated glucose value and approximately one hour of
historical trend data on its user interface, which is described in
more detail elsewhere herein.
[0226] FIG. 3C illustrates a third embodiment wherein the receiver
30 shows an estimated glucose value and approximately three hours
of historical trend data on its user interface, which is described
in more detail elsewhere herein.
[0227] FIG. 3D illustrates a fourth embodiment wherein the receiver
30 shows an estimated glucose value and approximately nine hours of
historical trend data on its user interface, which is described in
more detail elsewhere herein.
[0228] In some embodiments, a user can toggle through some or all
of the screens shown in FIGS. 3A to 3D using a toggle button on the
receiver. In some embodiments, the user will be able to
interactively select the type of output displayed on their user
interface. In other embodiments, the sensor output can have
alternative configurations.
[0229] FIG. 4A is a block diagram that illustrates one possible
configuration of the receiver's 30 electronics. It is noted that
the receiver 30 can comprise a configuration such as described with
reference to FIGS. 3A to 3D, above. Alternatively, the receiver 30
can comprise other configurations, including a desktop computer,
laptop computer, a personal digital assistant (PDA), a server
(local or remote to the receiver), and the like. In some
embodiments, the receiver 30 can be adapted to connect (via wired
or wireless connection) to a desktop computer, laptop computer,
PDA, server (local or remote to the receiver), and the like, in
order to download data from the receiver 30. In some alternative
embodiments, the receiver 30 and/or receiver electronics can be
housed within or directly connected to the sensor (e.g., 10 or 100)
in a manner that allows sensor and receiver electronics to work
directly together and/or share data processing resources.
Accordingly, the receiver's electronics can be generally referred
to as a "computer system."
[0230] A quartz crystal 40 is operatively connected to an RF
transceiver 41 that together function to receive and synchronize
data streams (e.g., raw data streams transmitted from the RF
transceiver). Once received, a processor 42 processes the signals,
such as described below.
[0231] The processor 42, also referred to as the processor module,
is the central control unit that performs the processing, such as
storing data, analyzing data streams, calibrating analyte sensor
data, estimating analyte values, comparing estimated analyte values
with time corresponding measured analyte values, analyzing a
variation of estimated analyte values, downloading data, and
controlling the user interface by providing analyte values,
prompts, messages, warnings, alarms, and the like. The processor
includes hardware and software that performs the processing
described herein, for example flash memory provides permanent or
semi-permanent storage of data, storing data such as sensor ID,
receiver ID, and programming to process data streams (for example,
programming for performing estimation and other algorithms
described elsewhere herein) and random access memory (RAM) stores
the system's cache memory and is helpful in data processing.
[0232] In one exemplary embodiment, the processor is a
microprocessor that provides the processing, such as calibration
algorithms stored within an EEPROM 43. The EEPROM 43 is operatively
connected to the processor 42 and provides semi-permanent storage
of data, storing data such as receiver ID and programming to
process data streams (e.g., programming for performing calibration
and other algorithms described elsewhere herein). In this exemplary
embodiment, an SRAM 44 is used for the system's cache memory and is
helpful in data processing.
[0233] A battery 45 is operatively connected to the processor 42
and provides power for the receiver. In one embodiment, the battery
is a standard AAA alkaline battery, however any appropriately sized
and powered battery can be used. In some embodiments, a plurality
of batteries can be used to power the system. A quartz crystal 46
is operatively connected to the processor 42 and maintains system
time for the computer system as a whole.
[0234] A user interface 47 comprises a keyboard 2, speaker 3,
vibrator 4, backlight 5, liquid crystal display (LCD 6), and one or
more buttons 7. The components that comprise the user interface 47
provide controls to interact with the user. The keyboard 2 can
allow, for example, input of user information about
himself/herself, such as mealtime, exercise, insulin
administration, and reference glucose values. The speaker 3 can
provide, for example, audible signals or alerts for conditions such
as present and/or predicted hyper- and hypoglycemic conditions. The
vibrator 4 can provide, for example, tactile signals or alerts for
reasons such as described with reference to the speaker, above. The
backlight 5 can be provided, for example, to aid the user in
reading the LCD in low light conditions. The LCD 6 can be provided,
for example, to provide the user with visual data output such as is
illustrated in FIGS. 3A to 3D. The buttons 7 can provide for
toggle, menu selection, option selection, mode selection, and
reset, for example.
[0235] In some embodiments, prompts or messages can be displayed on
the user interface to convey information to the user, such as
reference outlier values, requests for reference analyte values,
therapy recommendations, deviation of the measured analyte values
from the estimated analyte values, and the like. Additionally,
prompts can be displayed to guide the user through calibration or
trouble-shooting of the calibration.
[0236] Input and Output
[0237] In general, the above-described estimative algorithms,
including estimation of measured analyte values and variation
analysis of the estimated analyte values are useful when provided
to a patient, doctor, family member, and the like. Even more, the
estimative algorithms are useful when they are able to provide
information helpful in modifying a patient's behavior so that they
experience less clinically risky situations and higher quality of
life than may otherwise be possible. Therefore, the above-described
data analysis can be output in a variety of forms useful in caring
for the health of a patient.
[0238] Output can be provided via a user interface, including but
not limited to, visually on a screen, audibly through a speaker, or
tactilely through a vibrator. Additionally, output can be provided
via wired or wireless connection to an external device, including
but not limited to, computer, laptop, server, personal digital
assistant, modem connection, insulin delivery mechanism, medical
device, or other device that can be useful in interfacing with the
receiver.
[0239] Output can be continuously provided, or certain output can
be selectively provided based on events, analyte concentrations and
the like. For example, an estimated analyte path can be
continuously provided to a patient on an LCD screen, while audible
alerts can be provided only during a time of existing or
approaching clinical risk to a patient. As another example,
estimation can be provided based on event triggers (for example,
when an analyte concentration is nearing or entering a clinically
risky zone). As yet another example, analyzed deviation of
estimated analyte values can be provided when a predetermined level
of variation (for example, due to known error or clinical risk) is
known.
[0240] In some embodiments, alarms prompt or alert a patient when a
measured or projected analyte value or rate of change simply passes
a predetermined threshold. In some embodiments, the clinical risk
alarms combine intelligent and dynamic estimative algorithms to
provide greater accuracy, more timeliness in pending danger,
avoidance of false alarms, and less annoyance for the patient. For
example, clinical risk alarms of these embodiments include dynamic
and intelligent estimative algorithms based on analyte value, rate
of change, acceleration, clinical risk, statistical probabilities,
known physiological constraints, and/or individual physiological
patterns, thereby providing more appropriate, clinically safe, and
patient-friendly alarms.
[0241] In some embodiments, clinical risk alarms can be activated
for a predetermined time period to allow for the user to attend to
his/her condition. Additionally, the clinical risk alarms can be
de-activated when leaving a clinical risk zone so as not to annoy
the patient by repeated clinical risk alarms, when the patient's
condition is improving.
[0242] In some embodiments, the dynamic and intelligent estimation
determines a possibility of the patient avoiding clinical risk,
based on the analyte concentration, the rate of change, and other
aspects of the dynamic and intelligent estimative algorithms of the
preferred embodiments. If there is minimal or no possibility of
avoiding the clinical risk, a clinical risk alarm will be
triggered. However, if there is a possibility of avoiding the
clinical risk, the system can wait a predetermined amount of time
and re-analyze the possibility of avoiding the clinical risk. In
some embodiments, when there is a possibility of avoiding the
clinical risk, the system will further provide targets, therapy
recommendations, or other information that can aid the patient in
proactively avoiding the clinical risk.
[0243] In some embodiments, a variety of different display methods
are used, such as described in the preferred embodiments, which can
be toggled through or selectively displayed to the user based on
conditions or by selecting a button, for example. As one example, a
simple screen can be normally shown that provides an overview of
analyte data, for example present analyte value and directional
trend. More complex screens can then be selected when a user
desires more detailed information, for example, historical analyte
data, alarms, clinical risk zones, and the like.
[0244] FIG. 4B is an illustration of the receiver in one embodiment
showing an analyte trend graph, including measured analyte values,
estimated analyte values, and a clinical risk zone. The receiver 30
includes an LCD screen 34, buttons 7, and a speaker 3 and/or
microphone. The screen 34 displays a trend graph in the form of a
line representing the historical trend of a patient's analyte
concentration. Although axes may or may not be shown on the screen
34, it is understood that a theoretical x-axis represents time and
a theoretical y-axis represents analyte concentration.
[0245] In some embodiments such as shown in FIG. 4B, the screen
shows thresholds, including a high threshold 200 and a low
threshold 202, which represent boundaries between clinically safe
and clinically risky conditions for the patients. In one exemplary
embodiment, a normal glucose threshold for a glucose sensor is set
between about 100 and 160 mg/dL, and the clinical risk zones 204
are illustrated outside of these thresholds. In alternative
embodiments, the normal glucose threshold is between about 80 and
about 200 mg/dL, between about 55 and about 220 mg/dL, or other
threshold that can be set by the manufacturer, physician, patient,
computer program, and the like. Although a few examples of glucose
thresholds are given for a glucose sensor, the setting of any
analyte threshold is not limited by the preferred embodiments.
[0246] In some embodiments, the screen 34 shows clinical risk zones
204, also referred to as danger zones, through shading, gradients,
or other graphical illustrations that indicate areas of increasing
clinical risk. Clinical risk zones 204 can be set by a
manufacturer, customized by a doctor, and/or set by a user via
buttons 7, for example. In some embodiments, the danger zone 204
can be continuously shown on the screen 34, or the danger zone can
appear when the measured and/or estimated analyte values fall into
the danger zone 204. Additional information can be displayed on the
screen, such as an estimated time to clinical risk. In some
embodiments, the danger zone can be divided into levels of danger
(for example, low, medium, and high) and/or can be color-coded (for
example, yellow, orange, and red) or otherwise illustrated to
indicate the level of danger to the patient. Additionally, the
screen or portion of the screen can dynamically change colors or
illustrations that represent a nearness to the clinical risk and/or
a severity of clinical risk.
[0247] In some embodiments, such as shown in FIG. 4B, the screen 34
displays a trend graph of measured analyte data 206. Measured
analyte data can be smoothed and calibrated such as described in
more detail elsewhere herein. Measured analyte data can be
displayed for a certain time period (for example, previous 1 hour,
3 hours, 9 hours, etc.) In some embodiments, the user can toggle
through screens using buttons 7 to view the measured analyte data
for different time periods, using different formats, or to view
certain analyte values (for example, highs and lows).
[0248] In some embodiments such as shown in FIG. 4B, the screen 34
displays estimated analyte data 208 using dots. In this
illustration, the size of the dots can represent the confidence of
the estimation, a variation of estimated values, and the like. For
example, as the time gets farther away from the present (t=0) the
confidence level in the accuracy of the estimation can decline as
is appreciated by one skilled in the art. In some alternative
embodiments, dashed lines, symbols, icons, and the like can be used
to represent the estimated analyte values. In some alternative
embodiments, shaded regions, colors, patterns, and the like can
also be used to represent the estimated analyte values, a
confidence in those values, and/or a variation of those values,
such as described in more detail in preferred embodiments.
[0249] Axes, including time and analyte concentration values, can
be provided on the screen, however are not required. While not
wishing to be bound by theory, it is believed that trend
information, thresholds, and danger zones provide sufficient
information to represent analyte concentration and clinically
educate the user. In some embodiments, time can be represented by
symbols, such as a sun and moon to represent day and night. In some
embodiments, the present or most recent measured analyte
concentration, from the continuous sensor and/or from the reference
analyte monitor can be continually, intermittently, or selectively
displayed on the screen.
[0250] The estimated analyte values 208 of FIG. 4B include a
portion, which extends into the danger zone 204. By providing data
in a format that emphasizes the possibility of clinical risk to the
patient, appropriate action can be taken by the user (for example,
patient, or caretaker) and clinical risk can be preempted.
[0251] FIG. 4C is an illustration of the receiver in another
embodiment showing a representation of analyte concentration and
directional trend using a gradient bar. In this embodiment, the
screen illustrates the measured analyte values and estimated
analyte values in a simple but effective manner that communicates
valuable analyte information to the user.
[0252] In this embodiment, a gradient bar 210 is provided that
includes thresholds 212 set at high and lows such as described in
more detail with reference to FIG. 4B, above. Additionally, colors,
shading, or other graphical illustration can be present to
represent danger zones 214 on the gradient bar 210 such as
described in more detail with reference to FIG. 4B, above.
[0253] The measured analyte value is represented on the gradient
bar 210 by a marker 216, such as a darkened or colored bar. By
representing the measured analyte value with a bar 216, a
low-resolution analyte value is presented to the user (for example,
within a range of values). For example, each segment on the
gradient bar 210 can represent about 10 mg/dL of glucose
concentration. As another example, each segment can dynamically
represent the range of values that fall within the "A" and "B"
regions of the Clarke Error Grid. While not wishing to be bound by
theory, it is believed that inaccuracies known both in reference
analyte monitors and/or continuous analyte sensors are likely due
to known variables such as described in more detail elsewhere
herein, and can be de-emphasized such that a user focuses on
proactive care of the condition, rather than inconsequential
discrepancies within and between reference analyte monitors and
continuous analyte sensors.
[0254] Additionally, the representative gradient bar communicates
the directional trend of the analyte concentration to the user in a
simple and effective manner, namely by a directional arrow 218. For
example, in conventional diabetic blood glucose monitoring, a
person with diabetes obtains a blood sample and measures the
glucose concentration using a test strip, and the like.
Unfortunately, this information does not tell the person with
diabetes whether the blood glucose concentration is rising or
falling. Rising or falling directional trend information can be
particularly important in a situation such as illustrated in FIG.
4C, wherein if the user does not know that the glucose
concentration is rising, he/she may assume that the glucose
concentration is falling and not attend to his/her condition.
However, because rising directional trend information 218 is
provided, the person with diabetes can preempt the clinical risk by
attending to his/her condition (for example, administer insulin).
Estimated analyte data can be incorporated into the directional
trend information by characteristics of the arrow, for example,
size, color, flash speed, and the like.
[0255] In some embodiments, the gradient bar can be a vertical
instead of horizontal bar. In some embodiments, a gradient fill can
be used to represent analyte concentration, variation, or clinical
risk, for example. In some embodiments, the bar graph includes
color, for example the center can be green in the safe zone that
graduates to red in the danger zones; this can be in addition to or
in place of the divided segments. In some embodiments, the segments
of the bar graph are clearly divided by lines; however color,
gradation, and the like can be used to represent areas of the bar
graph. In some embodiments, the directional arrow can be
represented by a cascading level of arrows to a represent slow or
rapid rate of change. In some embodiments, the directional arrow
can be flashing to represent movement or pending danger.
[0256] The screen 34 of FIG. 4C can further comprise a numerical
representation of analyte concentration, date, time, or other
information to be communicated to the patient. However, a user can
advantageously extrapolate information helpful for his/her
condition using the simple and effective representation of this
embodiment shown in FIG. 4C, without reading a numeric
representation of his/her analyte concentration.
[0257] In some alternative embodiments, a trend graph or gradient
bar, a dial, pie chart, or other visual representation can provide
analyte data using shading, colors, patterns, icons, animation, and
the like.
[0258] FIG. 4D is an illustration of a receiver 30 in another
embodiment, including a screen 34 that shows a numerical
representation of the most recent measured analyte value 252. This
numerical value 252 is preferably a calibrated analyte value, such
as described in more detail with reference to FIGS. 5 and 6.
Additionally, this embodiment preferably provides an arrow 254 on
the screen 34, which represents the rate of change of the host's
analyte concentration. A bold "up" arrow is shown on the drawing,
which preferably represents a relatively quickly increasing rate of
change. The arrows shown with dotted lines illustrate examples of
other directional arrows (for example, rotated by 45 degrees),
which can be useful on the screen to represent various other
positive and negative rates of change. Although the directional
arrows shown have a relative low resolution (45 degrees of
accuracy), other arrows can be rotated with a high resolution of
accuracy (for example one degree of accuracy) to more accurately
represent the rate of change of the host's analyte concentration.
In some alternative embodiments, the screen provides an indication
of the acceleration of the host's analyte concentration.
[0259] A second numerical value 256 is shown, which is
representative of a variation of the measured analyte value 252.
The second numerical value is preferably determined from a
variation analysis based on statistical, clinical, or physiological
parameters, such as described in more detail elsewhere herein. In
one embodiment, the second numerical value 256 is determined based
on clinical risk (for example, weighted for the greatest possible
clinical risk to a patient). In another embodiment, the second
numerical representation 256 is an estimated analyte value
extrapolated to compensate for a time lag, such as described in
more detail elsewhere herein. In some alternative embodiments, the
receiver displays a range of numerical analyte values that best
represents the host's estimated analyte value (for example,
+/-10%). In some embodiments, the range is weighted based on
clinical risk to the patient. In some embodiments, the range is
representative of a confidence in the estimated analyte value
and/or a variation of those values. In some embodiments, the range
is adjustable.
[0260] Referring again to FIG. 4A, communication ports, including a
PC communication (com) port 48 and a reference glucose monitor com
port 49 can be provided to enable communication with systems that
are separate from, or integral with, the receiver 30. The PC com
port 48, for example, a serial communications port, allows for
communicating with another computer system (e.g., PC, PDA, server,
and the like). In one exemplary embodiment, the receiver 30 is able
to download historical data to a physician's PC for retrospective
analysis by the physician. The reference glucose monitor com port
49 allows for communicating with a reference glucose monitor (not
shown) so that reference glucose values can be downloaded into the
receiver 30, for example, automatically. In one embodiment, the
reference glucose monitor is integral with the receiver 30, and the
reference glucose com port 49 allows internal communication between
the two integral systems. In another embodiment, the reference
glucose monitor com port 49 allows a wireless or wired connection
to reference glucose monitor such as a self-monitoring blood
glucose monitor (e.g., for measuring finger stick blood
samples).
[0261] Calibration
[0262] Reference is now made to FIG. 5, which is a flow chart 50
that illustrates the process of initial calibration and data output
of the glucose sensor (e.g., 10 or 100) in one embodiment.
[0263] Calibration of the glucose sensor comprises data processing
that converts a sensor data stream into an estimated glucose
measurement that is meaningful to a user. Accordingly, a reference
glucose value can be used to calibrate the data stream from the
glucose sensor. In one embodiment, the analyte sensor is a
continuous glucose sensor and one or more reference glucose values
are used to calibrate the data stream from the sensor. The
calibration can be performed on a real-time basis and/or
retrospectively recalibrated. However in alternative embodiments,
other calibration techniques can be utilized, for example using
another constant analyte (for example, folic acid, ascorbate,
urate, and the like) as a baseline, factory calibration, periodic
clinical calibration, oxygen calibration (for example, using a
plurality of sensor heads), and the like can be used.
[0264] At block 51, a sensor data receiving module, also referred
to as the sensor data module, or processor module, receives sensor
data (e.g., a data stream), including one or more time-spaced
sensor data points hereinafter referred to as "data stream,"
"sensor data," "sensor analyte data", "glucose signal," from a
sensor via the receiver, which can be in wired or wireless
communication with the sensor. The sensor data can be raw or
smoothed (filtered), or include both raw and smoothed data. In some
embodiments, raw sensor data may include an integrated digital data
value, e.g., a value averaged over a time period such as by a
charge capacitor. Smoothed sensor data point(s) can be filtered in
certain embodiments using a filter, for example, a finite impulse
response (FIR) or infinite impulse response (IIR) filter. Some or
all of the sensor data point(s) can be replaced by estimated signal
values to address signal noise such as described in more detail
elsewhere herein. It is noted that during the initialization of the
sensor, prior to initial calibration, the receiver 30 (e.g.,
computer system) receives and stores the sensor data, however it
may not display any data to the user until initial calibration and
eventually stabilization of the sensor has been determined.
[0265] At block 52, a reference data receiving module, also
referred to as the reference input module, or the processor module,
receives reference data from a reference glucose monitor, including
one or more reference data points. In one embodiment, the reference
glucose points can comprise results from a self-monitored blood
glucose test (e.g., from a finger stick test). In one such
embodiment, the user can administer a self-monitored blood glucose
test to obtain a glucose value (e.g., point) using any known
glucose sensor, and enter the numeric glucose value into the
computer system. In another such embodiment, a self-monitored blood
glucose test comprises a wired or wireless connection to the
receiver 30 (e.g. computer system) so that the user simply
initiates a connection between the two devices, and the reference
glucose data is passed or downloaded between the self-monitored
blood glucose test and the receiver 30. In yet another such
embodiment, the self-monitored glucose test is integral with the
receiver 30 so that the user simply provides a blood sample to the
receiver 30, and the receiver 30 runs the glucose test to determine
a reference glucose value.
[0266] In some embodiments, the calibration process 50 monitors the
continuous analyte sensor data stream to determine a preferred time
for capturing reference analyte concentration values for
calibration of the continuous sensor data stream. In an example
wherein the analyte sensor is a continuous glucose sensor, when
data (for example, observed from the data stream) changes too
rapidly, the reference glucose value may not be sufficiently
reliable for calibration due to unstable glucose changes in the
host. In contrast, when sensor glucose data are relatively stable
(for example, relatively low rate of change), a reference glucose
value can be taken for a reliable calibration. In one embodiment,
the calibration process 38 can prompt the user via the user
interface to "calibrate now" when the analyte sensor is considered
stable.
[0267] In some embodiments, the calibration process 50 can prompt
the user via the user interface 47 to obtain a reference analyte
value for calibration at intervals, for example when analyte
concentrations are at high and/or low values. In some additional
embodiments, the user interface 47 can prompt the user to obtain a
reference analyte value for calibration based upon certain events,
such as meals, exercise, large excursions in analyte levels, faulty
or interrupted data readings, and the like. In some embodiments,
the estimative algorithms can provide information useful in
determining when to request a reference analyte value. For example,
when estimated analyte values indicate approaching clinical risk,
the user interface 47 can prompt the user to obtain a reference
analyte value.
[0268] Certain acceptability parameters can be set for reference
values received from the user. For example, in one embodiment, the
receiver may only accept reference glucose values between about 40
and about 400 mg/dL.
[0269] In some embodiments, the calibration process 50 performs
outlier detection on the reference data and time corresponding
sensor data. Outlier detection compares a reference analyte value
with a time corresponding measured analyte value to ensure a
predetermined statistically, physiologically, or clinically
acceptable correlation between the corresponding data exists. In an
example wherein the analyte sensor is a glucose sensor, the
reference glucose data is matched with substantially time
corresponding calibrated sensor data and the matched data are
plotted on a Clarke Error Grid to determine whether the reference
analyte value is an outlier based on clinical acceptability, such
as described in more detail with reference U.S. Publication No.
US-2005-0027463-A1. In some embodiments, outlier detection compares
a reference analyte value with a corresponding estimated analyte
value, such as described in more detail elsewhere herein and with
reference to the above-described patent application, and the
matched data is evaluated using statistical, clinical, and/or
physiological parameters to determine the acceptability of the
matched data pair. In alternative embodiments, outlier detection
can be determined by other clinical, statistical, and/or
physiological boundaries.
[0270] In some embodiments, outlier detection utilizes signal
artifacts detection, described in more detail elsewhere herein, to
determine the reliability of the reference data and/or sensor data
responsive to the results of the signal artifacts detection. For
example, if a certain level of signal artifacts is not detected in
the data signal, then the sensor data is determined to be reliable.
As another example, if a certain level of signal artifacts are
detected in the data signal, then the reliability of the reference
glucose data if the signal artifact is determined.
[0271] At block 53, a data matching module, also referred to as the
processor module, matches reference data (e.g., one or more
reference glucose data points) with substantially time
corresponding sensor data (e.g., one or more sensor data points) to
provide one or more matched data pairs. In one embodiment, one
reference data point is matched to one time corresponding sensor
data point to form a matched data pair. In another embodiment, a
plurality of reference data points are averaged (e.g., equally or
non-equally weighted average, mean-value, median, and the like) and
matched to one time corresponding sensor data point to form a
matched data pair. In another embodiment, one reference data point
is matched to a plurality of time corresponding sensor data points
averaged to form a matched data pair. In yet another embodiment, a
plurality of reference data points are averaged and matched to a
plurality of time corresponding sensor data points averaged to form
a matched data pair.
[0272] In one embodiment, a time corresponding sensor data
comprises one or more sensor data points that occur, for example,
15.+-.5 min after the reference glucose data timestamp (e.g., the
time that the reference glucose data is obtained). In this
embodiment, the 15 minute time delay has been chosen to account for
an approximately 10 minute delay introduced by the filter used in
data smoothing and an approximately 5 minute diffusional time-lag
(e.g., the time necessary for the glucose to diffusion through a
membrane(s) of a glucose sensor). In alternative embodiments, the
time corresponding sensor value can be more or less than in the
above-described embodiment, for example .+-.60 minutes. Variability
in time correspondence of sensor and reference data can be
attributed to, for example, a longer or shorter time delay
introduced during signal estimation, or if the configuration of the
glucose sensor incurs a greater or lesser physiological time
lag.
[0273] In some practical implementations of the sensor, the
reference glucose data can be obtained at a time that is different
from the time that the data is input into the receiver 30.
Accordingly, it should be noted that the "time stamp" of the
reference glucose (e.g., the time at which the reference glucose
value was obtained) may not be the same as the time at which the
receiver 30 obtained the reference glucose data. Therefore, some
embodiments include a time stamp requirement that ensures that the
receiver 30 stores the accurate time stamp for each reference
glucose value, that is, the time at which the reference value was
actually obtained from the user.
[0274] In some embodiments, tests are used to evaluate the
best-matched pair using a reference data point against individual
sensor values over a predetermined time period (e.g., about 30
minutes). In one such embodiment, the reference data point is
matched with sensor data points at 5-minute intervals and each
matched pair is evaluated. The matched pair with the best
correlation can be selected as the matched pair for data
processing. In some alternative embodiments, matching a reference
data point with an average of a plurality of sensor data points
over a predetermined time period can be used to form a matched
pair.
[0275] In some embodiments wherein the data signal is evaluated for
signal artifacts, as described in more detail elsewhere herein, the
processor module is configured to form a matching data pair only if
a signal artifact is not detected. In some embodiments wherein the
data signal is evaluated for signal artifacts, the processor module
is configured to prompt a user for a reference glucose value during
a time when one or more signal artifact(s) is not detected.
[0276] At block 54, a calibration set module, also referred to as
the processor module, forms an initial calibration set from a set
of one or more matched data pairs, which are used to determine the
relationship between the reference glucose data and the sensor
glucose data, such as described in more detail with reference to
block 55, below.
[0277] The matched data pairs, which make up the initial
calibration set, can be selected according to predetermined
criteria. In some embodiments, the number (n) of data pair(s)
selected for the initial calibration set is one. In other
embodiments, n data pairs are selected for the initial calibration
set wherein n is a function of the frequency of the received
reference data points. In one exemplary embodiment, six data pairs
make up the initial calibration set. In another embodiment, the
calibration set includes only one data pair.
[0278] In some embodiments, the data pairs are selected only within
a certain glucose value threshold, for example wherein the
reference glucose value is between about 40 and about 400 mg/dL. In
some embodiments, the data pairs that form the initial calibration
set are selected according to their time stamp. In certain
embodiments, the data pairs that form the initial calibration set
are selected according to their time stamp, for example, by waiting
a predetermined "break-in" time period after implantation, the
stability of the sensor data can be increased. In certain
embodiments, the data pairs that form the initial calibration set
are spread out over a predetermined time period, for example, a
period of two hours or more. In certain embodiments, the data pairs
that form the initial calibration set are spread out over a
predetermined glucose range, for example, spread out over a range
of at least 90 mg/dL or more.
[0279] In some embodiments, wherein the data signal is evaluated
for signal artifacts, as described in more detail elsewhere herein,
the processor module is configured to utilize the reference data
for calibration of the glucose sensor only if a signal artifact is
not detected.
[0280] At block 55, the conversion function module, also referred
to as the processor module, uses the calibration set to create a
conversion function. The conversion function substantially defines
the relationship between the reference glucose data and the glucose
sensor data. A variety of known methods can be used with the
preferred embodiments to create the conversion function from the
calibration set. In one embodiment, wherein a plurality of matched
data points form the initial calibration set, a linear least
squares regression is performed on the initial calibration set such
as described in more detail with reference to FIG. 6.
[0281] At block 56, a sensor data transformation module, also
referred to as the processor module, uses the conversion function
to transform sensor data into substantially real-time glucose value
estimates, also referred to as calibrated data, or converted sensor
data, as sensor data is continuously (or intermittently) received
from the sensor. For example, the sensor data, which can be
provided to the receiver in "counts," is translated in to estimate
analyte value(s) in mg/dL. In other words, the offset value at any
given point in time can be subtracted from the raw value (e.g., in
counts) and divided by the slope to obtain the estimated glucose
value:
mg / dL = ( rawvalue - offset ) slope ##EQU00001##
[0282] In some alternative embodiments, the sensor and/or reference
glucose values are stored in a database for retrospective
analysis.
[0283] At block 57, an output module, also referred to as the
processor module, provides output to the user via the user
interface. The output is representative of the estimated glucose
value, which is determined by converting the sensor data into a
meaningful glucose value such as described in more detail with
reference to block 56, above. User output can be in the form of a
numeric estimated glucose value, an indication of directional trend
of glucose concentration, and/or a graphical representation of the
estimated glucose data over a period of time, for example. Other
representations of the estimated glucose values are also possible,
for example audio and tactile.
[0284] In one embodiment, such as shown in FIG. 3A, the estimated
glucose value is represented by a numeric value. In other exemplary
embodiments, such as shown in FIGS. 3B to 3D, the user interface
graphically represents the estimated glucose data trend over
predetermined a time period (e.g., one, three, and nine hours,
respectively). In alternative embodiments, other time periods can
be represented. In alternative embodiments, other time periods can
be represented. In alternative embodiments, pictures, animation,
charts, graphs, ranges of values, and numeric data can be
selectively displayed.
[0285] Accordingly, after initial calibration of the sensor,
real-time continuous glucose information can be displayed on the
user interface so that the user can regularly and proactively care
for his/her diabetic condition within the bounds set by his/her
physician.
[0286] In alternative embodiments, the conversion function is used
to predict glucose values at future points in time. These predicted
values can be used to alert the user of upcoming hypoglycemic or
hyperglycemic events. Additionally, predicted values can be used to
compensate for a time lag (e.g., 15 minute time lag such as
described elsewhere herein), if any, so that an estimated glucose
value displayed to the user represents the instant time, rather
than a time delayed estimated value.
[0287] In some embodiments, the substantially real-time estimated
glucose value, a predicted future estimated glucose value, a rate
of change, and/or a directional trend of the glucose concentration
is used to control the administration of a constituent to the user,
including an appropriate amount and time, in order to control an
aspect of the user's biological system. One such example is a
closed loop glucose sensor and insulin pump, wherein the glucose
data (e.g., estimated glucose value, rate of change, and/or
directional trend) from the glucose sensor is used to determine the
amount of insulin, and time of administration, that can be given to
a diabetic user to evade hyper- and hypoglycemic conditions.
[0288] FIG. 6 is a graph that illustrates one embodiment of a
regression performed on a calibration set to create a conversion
function such as described with reference to FIG. 5, block 55,
above. In this embodiment, a linear least squares regression is
performed on the initial calibration set. The x-axis represents
reference glucose data; the y-axis represents sensor data. The
graph pictorially illustrates regression of matched pairs 66 in the
calibration set. The regression calculates a slope 62 and an offset
64, for example, using the well-known slope-intercept equation
(y=mx+b), which defines the conversion function.
[0289] In alternative embodiments, other algorithms could be used
to determine the conversion function, for example forms of linear
and non-linear regression, for example fuzzy logic, neural
networks, piece-wise linear regression, polynomial fit, genetic
algorithms, and other pattern recognition and signal estimation
techniques.
[0290] In yet other alternative embodiments, the conversion
function can comprise two or more different optimal conversions
because an optimal conversion at any time is dependent on one or
more parameters, such as time of day, calories consumed, exercise,
or glucose concentration above or below a set threshold, for
example. In one such exemplary embodiment, the conversion function
is adapted for the estimated glucose concentration (e.g., high vs.
low). For example in an implantable glucose sensor it has been
observed that the cells surrounding the implant will consume at
least a small amount of glucose as it diffuses toward the glucose
sensor. Assuming the cells consume substantially the same amount of
glucose whether the glucose concentration is low or high, this
phenomenon will have a greater effect on the concentration of
glucose during low blood sugar episodes than the effect on the
concentration of glucose during relatively higher blood sugar
episodes. Accordingly, the conversion function can be adapted to
compensate for the sensitivity differences in blood sugar level. In
one implementation, the conversion function comprises two different
regression lines, wherein a first regression line is applied when
the estimated blood glucose concentration is at or below a certain
threshold (e.g., 150 mg/dL) and a second regression line is applied
when the estimated blood glucose concentration is at or above a
certain threshold (e.g., 150 mg/dL). In one alternative
implementation, a predetermined pivot of the regression line that
forms the conversion function can be applied when the estimated
blood is above or below a set threshold (e.g., 150 mg/dL), wherein
the pivot and threshold are determined from a retrospective
analysis of the performance of a conversion function and its
performance at a range of glucose concentrations. In another
implementation, the regression line that forms the conversion
function is pivoted about a point in order to comply with clinical
acceptability standards (e.g., Clarke Error Grid, Consensus Grid,
mean absolute relative difference, or other clinical cost
function). Although only a few example implementations are
described, other embodiments include numerous implementations
wherein the conversion function is adaptively applied based on one
or more parameters that can affect the sensitivity of the sensor
data over time.
[0291] Additional methods for processing sensor glucose data are
disclosed in U.S. Publication No. US-2005-0027463-A1. In view of
the above-described data processing, it should be obvious that
improving the accuracy of the data stream will be advantageous for
improving output of glucose sensor data. Accordingly, the following
description is related to improving data output by decreasing
signal artifacts on the raw data stream from the sensor. The data
smoothing methods of preferred embodiments can be employed in
conjunction with any sensor or monitor measuring levels of an
analyte in vivo, wherein the level of the analyte fluctuates over
time, including but not limited to such sensors as described in
U.S. Pat. No. 6,001,067 to Shults et al.; U.S. Patent Application
2003/0023317 to Brauker et al.; U.S. Pat. No. 6,212,416 to Ward et
al.; U.S. Pat. No. 6,119,028 to Schulman et al; U.S. Pat. No.
6,400,974 to Lesho; U.S. Pat. No. 6,595,919 to Berner et al.; U.S.
Pat. No. 6,141,573 to Kurnik et al.; U.S. Pat. No. 6,122,536 to Sun
et al.; European Patent Application EP 1153571 to Varall et al.;
U.S. Pat. No. 6,512,939 to Colvin et al.; U.S. Pat. No. 5,605,152
to Slate et al.; U.S. Pat. No. 4,431,004 to Bessman et al.; U.S.
Pat. No. 4,703,756 to Gough et al; U.S. Pat. No. 6,514,718 to
Heller et al; and U.S. patent to U.S. Pat. No. 5,985,129 to Gough
et al.
[0292] Signal Artifacts
[0293] Typically, a glucose sensor produces a data stream that is
indicative of the glucose concentration of a host, such as
described in more detail above. However, it is well known that the
above described glucose sensors includes only a few examples of an
abundance of glucose sensors that are able to provide raw data
output indicative of the concentration of glucose. Thus, it should
be understood that the systems and methods described herein,
including signal artifacts detection, signal artifacts replacement,
and other data processing, can be applied to a data stream obtained
from any glucose sensor.
[0294] Raw data streams typically have some amount of "system
noise," caused by unwanted electronic or diffusion-related noise
that degrades the quality of the signal and thus the data.
Accordingly, conventional glucose sensors are known to smooth raw
data using methods that filter out this system noise, and the like,
in order to improve the signal to noise ratio, and thus data
output. One example of a conventional data-smoothing algorithm
includes a finite impulse response filter (FIR), which is
particularly suited for reducing high-frequency noise (see Steil et
al. U.S. Pat. No. 6,558,351).
[0295] FIGS. 7A and 7B are graphs of raw data streams from an
implantable glucose sensor prior to data smoothing. FIG. 7A is a
graph that shows a raw data stream obtained from a glucose sensor
over an approximately 4 hour time span in one example. FIG. 7B is a
graph that shows a raw data stream obtained from a glucose sensor
over an approximately 36 hour time span in another example. The
x-axis represents time in minutes. The y-axis represents sensor
data in counts. In these examples, sensor output in counts is
transmitted every 30-seconds.
[0296] The "system noise" such as shown in sections 72a, 72b of the
data streams of FIGS. 7A and 7B, respectively, illustrate time
periods during which system noise can be seen on the data stream.
This system noise can be characterized as Gaussian, Brownian,
and/or linear noise, and can be substantially normally distributed
about the mean. The system noise is likely electronic and
diffusion-related, and the like, and can be smoothed using
techniques such as by using an FIR filter. As another example, the
raw data can be represented by an integrated value, for example, by
integrating the signal over a time period (e.g., 30 seconds or 5
minutes), and providing an averaged (e.g., integrated) data point
there from. The system noise such as shown in the data of sections
72a, 72b is a fairly accurate representation of glucose
concentration and can be confidently used to report glucose
concentration to the user when appropriately calibrated.
[0297] The "signal artifacts" such as shown in sections 74a, 74b of
the data stream of FIGS. 7A and 7B, respectively, illustrate time
periods during which "signal artifacts" can be seen, which are
significantly different from the previously described system noise
(sections 72a, 72b). This noise, such as shown in section 74a and
74b, is referred to herein as "signal artifacts" and may be
described as "transient non-glucose dependent signal artifacts that
have a higher amplitude than system noise." At times, signal
artifacts comprise low noise, which generally refers to noise that
substantially decreases signal amplitude 76a, 76b herein, which is
best seen in the signal artifacts 74b of FIG. 7B. Occasional high
spikes 78a, 78b, which generally correspond to noise that
substantially increases signal amplitude, can also be seen in the
signal artifacts, which generally occur after a period of low
noise. These high spikes are generally observed after transient low
noise and typically result after reaction rate-limiting phenomena
occur. For example, in an embodiment where a glucose sensor
requires an enzymatic reaction, local ischemia creates a reaction
that is rate-limited by oxygen, which is responsible for low noise.
In this situation, glucose would be expected to build up in the
membrane because it would not be completely catabolized during the
oxygen deficit. When oxygen is again in excess, there would also be
excess glucose due to the transient oxygen deficit. The enzyme rate
would speed up for a short period until the excess glucose is
catabolized, resulting in high noise. Additionally, noise can be
distributed both above and below the expected signal.
[0298] Analysis of signal artifacts such as shown sections 74a, 74b
of FIGS. 7A and 7B, respectively, indicates that the observed low
noise is caused by substantially non-glucose reaction dependent
phenomena, such as ischemia that occurs within or around a glucose
sensor in vivo, for example, which results in the reaction becoming
oxygen dependent. As a first example, at high glucose levels,
oxygen can become limiting to the enzymatic reaction, resulting in
a non-glucose dependent downward trend in the data (best seen in
FIG. 7B). As a second example, certain movements or postures taken
by the patient can cause transient downward noise as blood is
squeezed out of the capillaries resulting in local ischemia, and
causing non-glucose dependent low noise. Because excess oxygen
(relative to glucose) is necessary for proper sensor function,
transient ischemia can result in a loss of signal gain in the
sensor data. In this second example oxygen can also become
transiently limited due to contracture of tissues around the sensor
interface. This is similar to the blanching of skin that can be
observed when one puts pressure on it. Under such pressure,
transient ischemia can occur in both the epidermis and subcutaneous
tissue. Transient ischemia is common and well tolerated by
subcutaneous tissue.
[0299] In another example of non-glucose reaction rate-limiting
phenomena, skin temperature can vary dramatically, which can result
in thermally related erosion of the signal (e.g., temperature
changes between 32 and 39 degrees Celsius have been measured in
humans). In yet another embodiment, wherein the glucose sensor is
placed intravenously, increased impedance can result from the
sensor resting against wall of the blood vessel, for example,
producing this non-glucose reaction rate-limiting noise due to
oxygen deficiency.
[0300] Because signal artifacts are not mere system noise, but
rather are caused by known or unknown non-glucose related
mechanisms, methods used for conventional random noise filtration
produce data lower (or in some cases higher) than the actual blood
glucose levels due to the expansive nature of these signal
artifacts. To overcome this, the preferred embodiments provide
systems and methods for replacing at least some of the signal
artifacts by estimating glucose signal values.
[0301] FIG. 8 is a flow chart that illustrates the process of
detecting and replacing signal artifacts in certain embodiments. It
is noted that "signal artifacts" particularly refers to the
transient non-glucose related artifacts such as described in more
detail elsewhere herein. Typically, signal artifacts are caused by
non-glucose rate-limiting phenomenon such as described in more
detail above.
[0302] At block 82, a sensor data receiving module, also referred
to as the sensor data module 82, or processor module, receives
sensor data (e.g., a data stream), including one or more
time-spaced sensor data points. In some embodiments, the data
stream is stored in the sensor for additional processing; in some
alternative embodiments, the sensor periodically transmits the data
stream to the receiver 30, which can be in wired or wireless
communication with the sensor. In some embodiments, raw and/or
filtered data is stored in the sensor and/or receiver.
[0303] At block 84, a signal artifacts detection module, also
referred to as the signal artifacts detector 84 or signal
reliability module, is programmed to detect transient non-glucose
related signal artifacts in the data stream, such as described in
more detail with reference to FIGS. 7A and 7B, above. The signal
artifacts detector can comprise an oxygen detector, a pH detector,
a temperature detector, and/or a pressure/stress detector, for
example, the signal artifacts detector 29 in FIG. 2. In some
embodiments, the signal artifacts detector at block 84 is located
within the processor 22 in FIG. 2 and utilizes existing components
of the glucose sensor to detect signal artifacts, for example by
pulsed amperometric detection, counter electrode monitoring,
reference electrode monitoring, and frequency content monitoring,
which are described elsewhere herein. In yet other embodiments, the
data stream can be sent from the sensor to the receiver which
comprises programming in the processor 42 in FIG. 4 that performs
algorithms to detect signal artifacts, for example such as
described with reference to "Cone of Possibility Detection" method
and/or by comparing raw data vs. filtered data, both of which are
described in more detail below. Numerous embodiments for detecting
signal artifacts are described in more detail in the section
entitled, "Signal Artifacts Detection," all of which are
encompassed by the signal artifacts detection at block 84.
[0304] In certain embodiments, the processor module in either the
sensor electronics and/or the receiver electronics can evaluate an
intermittent or continuous signal-to-noise measurement to determine
aberrancy of sensor data responsive to a signal-to-noise ratio
above a set threshold. In certain embodiments, signal residuals
(e.g., by comparing raw and filtered data) can be intermittently or
continuously analyzed for noise above a set threshold. In certain
embodiments, pattern recognition can be used to identify noise
associated with physiological conditions, such as low oxygen, or
other known signal aberrancies. Accordingly, in these embodiments,
the system can be configured, in response to aberrancies in the
data stream, to trigger signal estimation, adaptively filter the
data stream according to the aberrancy, and the like, as described
in more detail elsewhere herein.
[0305] At block 86, the signal artifacts replacement module, also
referred to as the signal estimation module, replaces some or an
entire data stream with estimated glucose signal values using
signal estimation. Numerous embodiments for performing signal
estimation are described in more detail in the section entitled
"Signal Artifacts Replacement," all of which are encompassed by the
signal artifacts replacement module, block 86. It is noted that in
some embodiments, signal estimation/replacement is initiated in
response to positive detection of signal artifacts on the data
stream, and subsequently stopped in response to detection of
negligible signal artifacts on the data stream. In some
embodiments, the system waits a predetermined time period (e.g.,
between 30 seconds and 30 minutes) before switching the signal
estimation on or off to ensure that a consistent detection has been
ascertained. In some embodiments, however, signal
estimation/replacement can continuously or continually run.
[0306] Some embodiments of signal estimation can additionally
include discarding data that is considered sufficiently unreliable
and/or erroneous such that the data should not be used in a signal
estimation algorithm. In these embodiments, the system can be
programmed to discard outlier data points, for example data points
that are so extreme that they can skew the data even with the most
comprehensive filtering or signal estimation, and optionally
replace those points with a projected value based on historical
data or present data (e.g., linear regression, recursive filtering,
and the like). One example of discarding sensor data includes
discarding sensor data that falls outside of a "Cone of
Possibility" such as described in more detail elsewhere herein.
Another example includes discarding sensor data when signal
artifacts detection detects values outside of a predetermined
threshold (e.g., oxygen concentration below a set threshold,
temperature above a certain threshold, signal amplitude above a
certain threshold, etc). Any of the signal estimation/replacement
algorithms described herein can then be used to project data values
for those data that were discarded.
[0307] Signal Artifacts Detection
[0308] Analysis of signals from glucose sensors indicates at least
two types of noise, which are characterized herein as 1) system
noise and 2) signal artifacts, such as described in more detail
above. It is noted that system noise is easily smoothed using the
algorithms provided herein; however, the systems and methods
described herein particularly address signal artifacts, by
replacing transient erroneous signal noise caused by rate-limiting
phenomenon (e.g., non-glucose related signal) with estimated signal
values, for example.
[0309] In certain embodiments of signal artifacts detection, oxygen
monitoring is used to detect whether transient non-glucose
dependent signal artifacts due to ischemia. Low oxygen
concentrations in or near the glucose sensor can account for a
large part of the transient non-glucose related signal artifacts as
defined herein on a glucose sensor signal, particularly in
subcutaneously implantable glucose sensors. Accordingly, detecting
oxygen concentration, and determining if ischemia exists can
discover ischemia-related signal artifacts. A variety of methods
can be used to test for oxygen. For example, an oxygen-sensing
electrode, or other oxygen sensor can be employed. The measurement
of oxygen concentration can be sent to a processor, which
determines if the oxygen concentration indicates ischemia.
[0310] In some embodiments of ischemia detection, an oxygen sensor
is placed proximal to or within the glucose sensor. For example,
the oxygen sensor can be located on or near the glucose sensor such
that their respective local environments are shared and oxygen
concentration measurement from the oxygen sensor represents an
accurate measurement of the oxygen concentration on or within the
glucose sensor. In some alternative embodiments of ischemia
detection, an oxygen sensor is also placed distal to the glucose
sensor. For example, the oxygen sensor can be located sufficiently
far from the glucose sensor such that their respective local
environments are not shared and oxygen measurements from the
proximal and distal oxygen sensors can be compared to determine the
relative difference between the respective local environments. By
comparing oxygen concentration at proximal and distal oxygen
sensors, change in local (proximal) oxygen concentration can be
determined from a reference (distal) oxygen concentration.
[0311] Oxygen sensors are useful for a variety of purposes. For
example, U.S. Pat. No. 6,512,939 to Colvin et al., which is
incorporated herein by reference, discloses an oxygen sensor that
measures background oxygen levels. However, Colvin et al. rely on
the oxygen sensor for the data stream of glucose measurements by
subtraction of oxygen remaining after exhaustion of glucose by an
enzymatic reaction from total unreacted oxygen concentration.
[0312] In another embodiment of ischemia detection, wherein the
glucose sensor is an electrochemical sensor that includes a
potentiostat, pulsed amperometric detection can be employed to
determine an oxygen measurement. Pulsed amperometric detection
includes switching, cycling, or pulsing the voltage of the working
electrode (or reference electrode) in an electrochemical system,
for example between a positive voltage (e.g., +0.6 for detecting
glucose) and a negative voltage (e.g., -0.6 for detecting oxygen).
U.S. Pat. No. 4,680,268 to Clark, Jr., which is incorporated by
reference herein, describes pulsed amperometric detection. In
contrast to using signal replacement, Clark, Jr. addresses oxygen
deficiency by supplying additional oxygen to the enzymatic
reaction.
[0313] In another embodiment of ischemia detection, wherein the
glucose sensor is an electrochemical sensor and contains a
potentiostat, oxygen deficiency can be seen at the counter
electrode when insufficient oxygen is available for reduction,
which thereby affects the counter electrode in that it is unable to
balance the current coming from the working electrode. When
insufficient oxygen is available for the counter electrode, the
counter electrode can be driven in its electrochemical search for
electrons all the way to its most negative value, which could be
ground or 0.0V, which causes the reference to shift, reducing the
bias voltage such as described in more detail below. In other
words, a common result of ischemia will be seen as a drop off in
sensor current as a function of glucose concentration (e.g., lower
sensitivity). This happens because the working electrode no longer
oxidizes all of the H.sub.2O.sub.2 arriving at its surface because
of the reduced bias. In some extreme circumstances, an increase in
glucose can produce no increase in current or even a decrease in
current.
[0314] FIG. 9 is a graph that shows a comparison of sensor current
and counter-electrode voltage in a host over time. The x-axis
represents time in minutes. The first y-axis 91 represents sensor
counts from the working electrode and thus plots a raw sensor data
stream 92 for the glucose sensor over a period of time. The second
y-axis 93 represents counter-electrode voltage 94 in counts. The
graph illustrates the correlation between sensor data 92 and
counter-electrode voltage 94; particularly, that erroneous counter
electrode function 96 where the counter voltages drops
approximately to zero substantially coincides with transient
non-glucose related signal artifacts 98. In other words, when
counter-electrode voltage is at or near zero, sensor data includes
signal artifacts.
[0315] In another embodiment of ischemia detection, wherein the
glucose sensor is an electrochemical sensor and contains a two- or
three-cell electrochemical cell, signal artifacts are detected by
monitoring the reference electrode. This "reference drift
detection" embodiment takes advantage of the fact that the
reference electrode will vary or drift in order to maintain a
stable bias potential with the working electrode, such as described
in more detail herein. This "drifting" generally indicates
non-glucose reaction rate-limiting noise, for example due to
ischemia. It is noted that the following example describes an
embodiment wherein the sensor includes a working, reference, and
counter electrodes, such as described in more detail elsewhere
herein; however the principles of this embodiment are applicable to
a two-cell (e.g., anode and cathode) electrochemical cell as is
understood in the art.
[0316] FIG. 10A is a circuit diagram of a potentiostat that
controls a typical three-electrode system, which can be employed
with a glucose sensor such as described with reference to FIGS. 1
and 2. The potentiostat includes a working electrode 100, a
reference electrode 102, and a counter electrode 104. The voltage
applied to the working electrode is a constant value (e.g., +1.2V)
and the voltage applied to the reference electrode is also set at a
constant value (e.g., +0.6V) such that the potential (V.sub.BIAS)
applied between the working and reference electrodes is maintained
at a constant value (e.g., +0.6V). The counter electrode is
configured to have a constant current (equal to the current being
measured by the working electrode), which is accomplished by
varying the voltage at the counter electrode in order to balance
the current going through the working electrode 100 such that
current does not pass through the reference electrode 102. A
negative feedback loop 107 is constructed from an operational
amplifier (OP AMP), the reference electrode 102, the counter
electrode 104, and a reference potential, to maintain the reference
electrode at a constant voltage.
[0317] In practice, a glucose sensor of one embodiment comprises a
membrane that contains glucose oxidase that catalyzes the
conversion of oxygen and glucose to hydrogen peroxide and
gluconate, such as described with reference to FIGS. 1 and 2.
Therefore, for each glucose molecule metabolized there is a change
equivalent in molecular concentration in the co-reactant O.sub.2
and the product H.sub.2O.sub.2. Consequently, one can use an
electrode (e.g., working electrode 100) to monitor the
concentration-induced current change in either the co-reactant or
the product to determine glucose concentration.
[0318] One limitation of the electrochemistry is seen when
insufficient negative voltage is available to the counter electrode
104 to balance the working electrode 100. This limitation can occur
when insufficient oxygen is available to the counter electrode 104
for reduction, for example. When this limitation occurs, the
counter electrode can no longer vary its voltage to maintain a
balanced current with the working electrode and thus the negative
feedback loop 107 used to maintain the reference electrode is
compromised. Consequently, the reference electrode voltage will
change or "drift," altering the applied bias potential (i.e., the
potential applied between the working and reference electrodes),
thereby decreasing the applied bias potential. When this change in
applied bias potential occurs, the working electrode can produce
erroneous glucose measurements due to either increased or decreased
signal strength (I.sub.SENSE).
[0319] FIG. 10B a diagram referred to as Cyclic-Voltammetry (CV)
curve, wherein the x-axis represents the applied potential
(V.sub.BIAS) and the y-axis represents the signal strength of the
working electrode (I.sub.SENSE). A curve 108 illustrates an
expected CV curve when the potentiostat is functioning normally.
Typically, desired bias voltage can be set (e.g., V.sub.BIAS1) and
a resulting current will be sensed (e.g., I.sub.SENSE1). As the
voltage decreases (e.g., V.sub.BIAS2) due to reference voltage
drift, for example, a new resulting current is sensed (e.g.,
I.sub.SENSE2). Therefore, the change in bias is an indicator of
signal artifacts and can be used in signal estimation and to
replace the erroneous resulting signals. In addition to ischemia,
the local environment at the electrode surfaces can affect the CV
curve, for example, changes in pH, temperature, and other local
biochemical species can significantly alter the location of the CV
curve.
[0320] FIG. 10C is a CV curve that illustrates an alternative
embodiment of signal artifacts detection, wherein pH and/or
temperature can be monitoring using the CV curve and diagnosed to
detect transient non-glucose related signal artifacts. For example,
signal artifacts can be attributed to thermal changes and/or pH
changes in some embodiments because certain changes in pH and
temperature affect data from a glucose sensor that relies on an
enzymatic reaction to measure glucose. Signal artifacts caused by
pH changes, temperature changes, changes in available electrode
surface area, and other local biochemical species can be detected
and signal estimation can be applied an/or optimized such as
described in more detail elsewhere herein. In FIG. 10C, a first
curve 108 illustrates an expected CV curve when the potentiostat is
functioning normally. A second curve 109 illustrates a CV curve
wherein the environment has changed as indicated by the upward
shift of the CV curve.
[0321] In some embodiments, pH and/or temperature measurements are
obtained proximal to the glucose sensor; in some embodiments, pH
and/or temperature measurements are also obtained distal to the
glucose sensor and the respective measurements compared, such as
described in more detail above with reference to oxygen
sensors.
[0322] In another implementation of signal artifacts detection,
wherein temperature is detected, an electronic thermometer can be
proximal to or within the glucose sensor, such that the temperature
measurement is representative of the temperature of the glucose
sensor's local environment. It is noted that accurate sensor
function depends on diffusion of molecules from the blood to the
interstitial fluid, and then through the membranes of the device to
the enzyme membrane. Additionally, diffusion transport of hydrogen
peroxide from the enzyme membrane to the electrode is required for
accurate sensor function in some embodiments. Therefore,
temperatures can be a rate determining parameter of diffusion. As
temperature decreases, diffusion transport decreases. Under certain
human conditions, such as hypothermia or fever, the variations can
be considerably greater. Additionally, under normal conditions, the
temperature of subcutaneous tissue is known to vary considerably
more than core tissues (e.g., core temperature). Temperature
thresholds can be set to detect signal artifacts accordingly.
[0323] In another implementation, a pH detector is used to detect
signal artifacts. In glucose sensors that rely on enzymatic
reactions, a pH of the fluid to be sensed can be within the range
of about 5.5 to 7.5. Outside of this range, effects may be seen in
the enzymatic reaction and therefore data output of the glucose
sensor. Accordingly, by detecting if the pH is outside of a
predetermined range (e.g., 5.5 to 7.5), a pH detector may detect
transient non-glucose related signal artifacts such as described
herein. It is noted that the pH threshold can be set at ranges
other than provided herein without departing from the preferred
embodiments.
[0324] In an alternative embodiment of signal artifacts detection,
pressure and/or stress can be monitored using known techniques for
example by a strain gauge placed on the sensor that detects
stress/strain on the circuit board, sensor housing, or other
components. A variety of microelectromechanical systems (MEMS) can
be utilized to measure pressure and/or stress within the
sensor.
[0325] In another alternative embodiment of signal artifacts
detection, the processor in the sensor (or receiver) periodically
evaluates the data stream for high amplitude noise, which is
defined by noisy data wherein the amplitude of the noise is above a
predetermined threshold. For example, in the graph of FIGS. 7A and
7B, the system noise sections such as 72a and 72b have a
substantially low amplitude noise threshold; in contrast to system
noise, signal artifacts sections such as 74a and 74b have signal
artifacts (noise) with an amplitude that is much higher than that
of system noise. Therefore, a threshold can be set at or above the
amplitude of system noise, such that when noisy data is detected
above that amplitude, it can be considered "signal artifacts" as
defined herein.
[0326] In another alternative embodiment of signal artifacts
detection, a method hereinafter referred to as the "Cone of
Possibility Detection Method," utilizes physiological information
along with glucose signal values in order define a "cone" of
physiologically feasible glucose signal values within a human, such
that signal artifacts are detected whenever the glucose signal
falls outside of the cone of possibility. Particularly,
physiological information depends upon the physiological parameters
obtained from continuous studies in the literature as well as our
own observations. A first physiological parameter uses a maximal
sustained rate of change of glucose in humans (e.g., about 4 to 5
mg/dL/min) and a maximum acceleration of that rate of change (e.g.,
about 0.1 to 0.2 mg/dL/min.sup.2). A second physiological parameter
uses the knowledge that rate of change of glucose is lowest at the
minima, which is the areas of greatest risk in patient treatment,
and the maxima, which has the greatest long-term effect on
secondary complications associated with diabetes. A third
physiological parameter uses the fact that the best solution for
the shape of the curve at any point along the curve over a certain
time period (e.g., about 20-30 minutes) is a straight line.
Additional physiological parameters can be incorporated and are
within the scope of this embodiment.
[0327] In practice, the Cone of Possibility Detection Method
combines any one or more of the above-described physiological
parameters to form an algorithm that defines a cone of possible
glucose levels for glucose data captured over a predetermined time
period. In one exemplary implementation of the Cone of Possibility
Detection Method, the system (processor in the sensor or receiver)
calculates a maximum physiological rate of change and determines if
the data falls within these physiological limits; if not, signal
artifacts are identified. It is noted that the maximum rate of
change can be narrowed (e.g., decreased) in some instances.
Therefore, additional physiological data could be used to modify
the limits imposed upon the Cone of Possibilities Detection Method
for sensor glucose values. For example, the maximum per minute rate
change can be lower when the subject is sleeping or hasn't eaten in
eight hours; on the other hand, the maximum per minute rate change
can be higher when the subject is exercising or has consumed high
levels of glucose, for example. In general, it has been observed
that rates of change are slowest near the maxima and minima of the
curve, and that rates of change are highest near the midpoint
between the maxima and minima. It should further be noted that rate
of change limits are derived from analysis of a range of data
significantly higher unsustained rates of change can be
observed.
[0328] In another alternative embodiment of signal artifacts
detection, examination of the spectral content (e.g., frequency
content) of the data stream can yield measures useful in detecting
signal artifacts. For example, data that has high frequency, and in
some cases can be characterized by a large negative slope, are
indicative of signal artifacts and can cause sensor signal loss.
Specific signal content can be monitored using an orthogonal
transform, for example a Fourier transform, a Discrete Fourier
Transform (DFT), or any other method known in the art.
[0329] FIG. 11 is a graph of 110 a raw data stream from a glucose
sensor and a spectrogram 114 that shows the frequency content of
the raw data stream in one embodiment. Particularly, the graph 110
illustrates the raw data stream 112 and includes an x-axis that
represents time in hours and a y-axis that represents sensor data
output in counts; the spectrogram 114 illustrates the frequency
content 116 corresponding to the raw data stream 112 and includes
an x-axis that represents time in hours corresponding to the x-axis
of the graph 110 and a y-axis that represents frequency content in
cycles per hour. The darkness of each point represents the
amplitude of that frequency at that time. Darker points relate to
higher amplitudes. Frequency content on the spectrogram 114 was
obtained using a windowed Discrete Fourier transform.
[0330] The raw data stream in the graph 110 has been adjusted by a
linear mapping similar to the calibration algorithm. In this
example, the bias (or intercept) has been adjusted but not the
proportion (or slope). The slope of the raw data stream would
typically be determined by some calibration, but since the
calibration has not occurred in this example, the gray levels in
the spectrogram 114 indicate relative values. The lower values of
the graph 110 are white. They are colored as white below a specific
value, highlighting only the most intense areas of the graph.
[0331] By monitoring the frequency content 116, high frequency
cycles 118 can be observed. The high frequency cycles 118
correspond to signal artifacts 119 such as described herein. Thus,
signal artifacts can be detected on a data stream by monitoring
frequency content and setting a threshold (e.g., 30 cycles per
hour). The set threshold can vary depending on the signal
source.
[0332] In another alternative embodiment of signal artifacts
detection, examination of the signal information content can yield
measures useful in detecting signal artifacts. Time series analysis
can be used to measure entropy, approximate entropy, variance,
and/or percent change of the information content over consecutive
windows (e.g., 30 and 60 minute windows of data) of the raw data
stream. In one exemplary embodiment, the variance of the raw data
signal is measured over 15 minute and 45 minute windows, and signal
artifacts are detected when the variance of the data within the
15-minute window exceeds the variance of the data within the
45-minute window. Alternatively, other methods of self-diagnosis
can be performed on the signal to determine a level of signal
artifacts. One example includes performing a first derivative
analysis that compares consecutive points, and detects signal
artifacts when point to point changes are above a physiologically
feasible threshold, for example. Another example of signal
self-diagnosis includes performing a second derivative analysis
that considers turning points, for example, detects signal
artifacts when changes are not sufficiently gradual (e.g., within
thresholds), for example.
[0333] In yet another alternative embodiment of signal artifacts
detection that utilizes examination or evaluation of the signal
information content, filtered (e.g., smoothed) data is compared to
raw data (e.g., in sensor electronics or in receiver electronics).
In one such embodiment, a signal residual is calculated as the
difference between the filtered data and the raw data. For example,
at one time point (or one time period that is represented by a
single raw value and single filtered value), the filtered data can
be measured at 50,000 counts and the raw data can be measured at
55,500 counts, which would result in a signal residual of 5,500
counts. In some embodiments, a threshold can be set (e.g., 5000
counts) that represents a first level of noise (e.g., signal
artifact) in the data signal, when the residual exceeds that level.
Similarly, a second threshold can be set (e.g., 8,000 counts) that
represents a second level of noise in the data signal. Additional
thresholds and/or noise classifications can be defined as is
appreciated by one skilled in the art. Consequently, signal
filtering, processing, and/or displaying decisions can be executed
based on these conditions (e.g., the predetermined levels of
noise).
[0334] Although the above-described example illustrates one method
of determining a level of noise, or signal artifact(s), based on a
comparison of raw vs. filtered data for a time point (or single
values representative of a time period). In an alternative
exemplary embodiment for determining noise, signal artifacts are
evaluated for noise episodes lasting a certain period of time. For
example, the processor (in the sensor or receiver) can be
configured to look for a certain number of signal residuals above a
predetermined threshold (representing noise time points or noisy
time periods) for a predetermined period of time (e.g., a few
minutes to a few hours or more).
[0335] In one exemplary embodiment, a processor is configured to
determine a signal residual by subtracting the filtered signal from
the raw signal for a predetermined time period. It is noted that
the filtered signal can be filtered by any known smoothing
algorithm such as described herein, for example a 3-point moving
average-type filter. It is further noted that the raw signal can
include an average value, e.g., wherein the value is integrated
over a predetermined time period (such as 5-minutes). Furthermore,
it is noted that the predetermined time period can be a time point
or representative data for a time period (e.g., 5 minutes). In some
embodiments, wherein a noise episode for a predetermined time
period is being evaluated, a differential can be obtained by
comparing a signal residual with a previous signal residual (e.g.,
a residual at time (t)=0 as compared to a residual at (t)-5
minutes.) Similar to the thresholds described above with regard to
the signal residual, one or more thresholds can be set for the
differentials, whereby one or more differentials above one of the
predetermined differential thresholds defines a particular noise
level. It has been shown in certain circumstances that a
differential measurement as compared to a residual measurement as
described herein, amplifies noise and therefore may be a more
sensitive to noise episodes. Accordingly, a noise episode, or noise
episode level, can be defined by one or more points (e.g.,
residuals or differentials) above a predetermined threshold, and in
some embodiments, for a predetermined period of time. Similarly, a
noise level determination can be reduced or altered when a
different (e.g., reduced) number of points above the predetermined
threshold are calculated in a predetermined period of time.
[0336] One or a plurality of the above signal artifacts detection
models can be used alone or in combination to detect signal
artifacts such as described herein. Accordingly, the data stream
associated with the signal artifacts can be discarded, replaced, or
otherwise processed in order to reduce or eliminate these signal
artifacts and thereby improve the value of the glucose measurements
that can be provided to a user.
[0337] Signal Artifacts Replacement
[0338] Signal Artifacts Replacement, such as described above, can
use systems and methods that reduce or replace these signal
artifacts that can be characterized by transience, high frequency,
high amplitude, and/or substantially non-linear noise. Accordingly,
a variety of filters, algorithms, and other data processing are
provided that address the detected signal artifacts by replacing
the data stream, or portion of the data stream, with estimated
glucose signal values. It is noted that "signal estimation" as used
herein, is a broad term, which includes filtering, data smoothing,
augmenting, projecting, and/or other algorithmic methods that
estimate glucose signal values based on present and historical
data.
[0339] It is noted that a glucose sensor can contain a processor
and the like that processes periodically received raw sensor data
(e.g., every 30 seconds). Although a data point can be available
constantly, for example by use of an electrical integration system
in a chemo-electric sensor, relatively frequent (e.g., every 30
seconds), or less frequent data point (e.g., every 5 minutes), can
be more than sufficient for patient use. It is noted that
accordingly Nyquist Theory, a data point is required about every 10
minutes to accurately describe physiological change in glucose in
humans. This represents the lowest useful frequency of sampling.
However, it should be recognized that it is desirable to sample
more frequently than the Nyquist minimum, to provide for sufficient
data in the event that one or more data points are lost, for
example. Additionally, more frequently sampled data (e.g.,
30-second) can be used to smooth the less frequent data (e.g.,
5-minute) that are transmitted. It is noted that in this example,
during the course of a 5-minute period, 10 determinations are made
at 30-second intervals.
[0340] In some embodiments of Signal Artifacts Replacement, signal
estimation can be implemented in the sensor and transmitted to a
receiver for additional processing. In some embodiments of Signal
Artifacts Replacement, raw data can be sent from the sensor to a
receiver for signal estimation and additional processing therein.
In some embodiments of Signal Artifacts Replacement, signal
estimation is performed initially in the sensor, with additional
signal estimation in the receiver.
[0341] In some embodiments of Signal Artifacts Replacement, wherein
the sensor is an implantable glucose sensor, signal estimation can
be performed in the sensor to ensure a continuous stream of data.
In alternative embodiments, data can be transmitted from the sensor
to the receiver, and the estimation performed at the receiver; It
is noted however that there can be a risk of transmit-loss in the
radio transmission from the sensor to the receiver when the
transmission is wireless. For example, in embodiments wherein a
sensor is implemented in vivo, the raw sensor signal can be more
consistent within the sensor (in vivo) than the raw signal
transmitted to a source (e.g., receiver) outside the body (e.g., if
a patient were to take the receiver off to shower, communication
between the sensor and receiver can be lost and data smoothing in
the receiver would halt accordingly). Consequently, It is noted
that a multiple point data loss in the filter can take for example,
about 25 to about 40 minutes for the data to recover to near where
it would have been had there been no data loss.
[0342] In some embodiments of Signal Artifacts Replacement, signal
estimation is initiated only after signal artifacts are positively
detected and stopped once signal artifacts are negligibly detected.
In some alternative embodiments signal estimation is initiated
after signal artifacts are positively detected and then stopped
after a predetermined time period. In some alternative embodiments,
signal estimation can be continuously or continually performed. In
some alternative embodiments, one or more forms of signal
estimation can be accomplished based on the severity of the signal
artifacts, such as will be described with reference the section
entitled, "Selective Application of Signal Artifacts
Replacement."
[0343] In some embodiments of Signal Artifacts Replacement, the
processor performs a linear regression. In one such implementation,
the processor performs a linear regression analysis of the n (e.g.,
10) most recent sampled sensor values to smooth out the noise. A
linear regression averages over a number of points in the time
course and thus reduces the influence of wide excursions of any
point from the regression line. Linear regression defines a slope
and intercept, which is used to generate a "Projected Glucose
Value," which can be used to replace sensor data. This regression
can be continually performed on the data stream or continually
performed only during the transient signal artifacts. In some
alternative embodiments, signal estimation can include non-linear
regression.
[0344] In another embodiment of Signal Artifacts Replacement, the
processor performs a trimmed regression, which is a linear
regression of a trimmed mean (e.g., after rejecting wide excursions
of any point from the regression line). In this embodiment, after
the sensor records glucose measurements at a predetermined sampling
rate (e.g., every 30 seconds), the sensor calculates a trimmed mean
(e.g., removes highest and lowest measurements from a data set and
then regresses the remaining measurements to estimate the glucose
value.
[0345] FIG. 12 is a graph that illustrates a raw data stream from a
glucose sensor and a trimmed regression that can be used to replace
some of or the entire data stream. The x-axis represents time in
minutes; the y-axis represents sensor data output in counts. A raw
data signal 120, which is illustrated as a dotted line, shows a
data stream wherein some system noise can be detected, however
signal artifacts 122 can be particularly seen in a portion thereof
(and can be detected by methods such as described above). The
trimmed regression line 124, which is illustrated as a solid line,
is the data stream after signal estimation using a trimmed linear
regression algorithm, such as described above, and appears at least
somewhat "smoothed" on the graph. In this particular example, the
trimmed regression uses the most recent 60 points (30 minutes) and
trims out the highest and lowest values, then uses the leftover 58
points to project the next point. It is noted that the trimmed
regression 124 provides a good estimate throughout the majority
data stream, however is only somewhat effective in smoothing the
data in during signal artifacts 122. To provide an optimized
estimate of the glucose data values, the trimmed regression can be
optimized by changing the parameters of the algorithm, for example
by trimming more or less raw glucose data from the top and/or
bottom of the signal artifacts 122 prior to regression.
Additionally, trimmed regression, because of its inherent
properties, can be particularly suited for noise of a certain
amplitude and/or characteristic. In one embodiment, for example
trimmed regression can be selectively applied based on the severity
of the signal artifacts, which is described in more detail below
with reference to FIGS. 15 to 17.
[0346] In another embodiment of Signal Artifacts Replacement, the
processor runs a non-recursive filter, such as a finite impulse
response (FIR) filter. A FIR filter is a digital signal filter, in
which every sample of output is the weighted sum of past and
current samples of input, using only some finite number of past
samples.
[0347] FIG. 13 is a graph that illustrates a raw data stream from a
glucose sensor and an FIR-estimated signal that can be used to
replace some of or the entire data stream. The x-axis represents
time in minutes; the y-axis represents sensor data output in
counts. A raw data signal 130, which is illustrated as a dotted
line, shows a data stream wherein some system noise can be
detected, however signal artifacts 132 can be particularly seen in
a portion thereof (and can be detected by methods such as described
above). The FIR-estimated signal 134, which is illustrated as a
solid line, is the data stream after signal estimation using a FIR
filter, such as described above, and appears at least somewhat
"smoothed" on the graph. It is noted that the FIR-estimated signal
provides a good estimate throughout the majority of the data
stream; however like trimmed regression it is only somewhat
effective in smoothing the data during signal artifacts 132. To
provide an optimized estimate of the glucose data values, the FIR
filter can be optimized by changing the parameters of the
algorithm, for example the tuning of the filter, particularly the
frequencies of the pass band and the stop band. Additionally, it is
noted that the FIR filter, because of its inherent properties, can
be particularly suited for noise of a certain amplitude and/or
characteristic. In one embodiment, for example the FIR filter can
be selectively applied based on the severity of the signal
artifacts, which is described in more detail below with reference
to FIGS. 15 to 17. It is noted that the FIR-estimated signal
induces a time lag on the data stream, which can be increased or
decreased in order to optimize the filtering or to minimize the
time lag, for example.
[0348] In another embodiment of Signal Artifacts Replacement, the
processor runs a recursive filter, such as an infinite impulse
response (IIR) filter. An IIR filter is a type of digital signal
filter, in which every sample of output is the weighted sum of past
and current samples of input. In one exemplary implementation of an
IIR filter, the output is computed using 6 additions/subtractions
and 7 multiplications as shown in the following equation:
y ( n ) = a 0 x ( n ) + a 1 x ( n - 1 ) + a 2 x ( n - 2 ) + a 3 x (
n - 3 ) - b 1 y ( n - 1 ) - b 2 y ( n - 2 ) - b 3 y ( n - 3 ) b 0
##EQU00002##
[0349] This polynomial equation includes coefficients that are
dependent on sample rate and frequency behavior of the filter.
Frequency behavior passes low frequencies up to cycle lengths of 40
minutes, and is based on a 30 second sample rate. In alternative
implementations, the sample rate and cycle lengths can be more or
less. See Lynn "Recursive Digital Filters for Biological Signals"
Med. & Biol. Engineering, Vol. 9, pp. 37-43, which is
incorporated herein by reference in its entirety.
[0350] FIG. 14 is a graph that illustrates a raw data stream from a
glucose sensor and an IIR-estimated signal that can be used to
replace some of or the entire data stream. The x-axis represents
time in minutes; the y-axis represents sensor data output in
counts. A raw data signal 140, which is illustrated as a dotted
line, shows a data stream wherein some system noise can be
detected, however signal artifacts 142 can be particularly seen in
a portion thereof (and can be detected by methods such as described
above). The IIR-estimated signal 144, which is illustrated as a
solid line, represents the data stream after signal estimation
using an IIR filter, such as described above, and appears at least
somewhat "smoothed" on the graph. It is noted that the
IIR-estimated signal induces a time lag on the data stream; however
it appears to be a particularly good estimate of glucose data
values during signal artifacts 142, as compared to the FIR filter
(FIG. 13), for example.
[0351] To optimize the estimation of the glucose data values, the
parameters of the IIR filter can be optimized, for example by
increasing or decreasing the cycle lengths (e.g., 10 minutes, 20
minute, 40 minutes, 60 minutes) that are used in the algorithm.
Although an increased cycle length can increase the time lag
induced by the IIR filter, an increased cycle length can also
better estimate glucose data values during severe signal artifacts.
In other words, the IIR filter, because of its inherent properties,
can be particularly suited for noise of a certain amplitude and/or
characteristic. In one exemplary embodiment, the IIR filter can be
continually applied, however the parameters such as described above
can be selectively altered based on the severity of the signal
artifacts; in another exemplary embodiment, the IIR filter can be
applied only after positive detection of signal artifacts.
Selective application of the IIR filter based on the severity of
the signal artifacts is described in more detail below with
reference to FIGS. 15 to 17.
[0352] It is noted that a comparison of linear regression, an FIR
filter, and an IIR filter can be advantageous for optimizing their
usage in the preferred embodiments. That is, an understanding the
design considerations for each algorithm can lead to optimized
selection and implementation of the algorithm, as described in the
section entitled, "Selective Application of Signal Replacement
Algorithms" herein. During system noise, as defined herein, all of
the above algorithms can be successfully implemented during system
noise with relative ease. During signal artifacts, however,
computational efficiency is greater with an IIR-filter as compared
with linear regression and FIR-filter. Additionally, although the
time lag associated with an IIR filter can be substantially greater
than that of the linear regression or FIR-filter, this can be
advantageous during severe signal artifacts in order to assign
greater weight toward the previous, less noisy data in signal
estimation.
[0353] In another embodiment of Signal Artifacts Replacement, the
processor runs a maximum-average (max-average) filtering algorithm.
The max-average algorithm smoothes data based on the discovery that
the substantial majority of signal artifacts observed after
implantation of glucose sensors in humans, for example, is not
distributed evenly above and below the actual blood glucose levels.
It has been observed that many data sets are actually characterized
by extended periods in which the noise appears to trend downwardly
from maximum values with occasional high spikes such as described
in more detail above with reference to FIG. 7B, section 74b, which
is likely in response to limitations in the system that do not
allow he glucose to fully react at the enzyme layer and/or proper
reduction of H.sub.2O.sub.2 at the counter electrode, for example.
To overcome these downward trending signal artifacts, the
max-average calculation tracks with the highest sensor values, and
discards the bulk of the lower values. Additionally, the
max-average method is designed to reduce the contamination of the
data with non-physiologically high data from the high spikes.
[0354] The max-average calculation smoothes data at a sampling
interval (e.g., every 30 seconds) for transmission to the receiver
at a less frequent transmission interval (e.g., every 5 minutes) to
minimize the effects of low non-physiological data. First, the
processor finds and stores a maximum sensor counts value in a first
set of sampled data points (e.g., 5 consecutive, accepted,
thirty-second data points). A frame shift time window finds a
maximum sensor counts value for each set of sampled data (e.g.,
each 5-point cycle length) and stores each maximum value. The
processor then computes a rolling average (e.g., 5-point average)
of these maxima for each sampling interval (e.g., every 30 seconds)
and stores these data. Periodically (e.g., every 10.sup.th
interval), the sensor outputs to the receiver the current maximum
of the rolling average (e.g., over the last 10 thirty-second
intervals as a smoothed value for that time period (e.g., 5
minutes)). In one example implementation, a 10-point window can be
used, and at the 10.sup.th interval, the processor computes the
average of the most recent 5 or 10 average maxima as the smoothed
value for a 5 minute time period.
[0355] In some embodiments of the max-average algorithm, an
acceptance filter can also be applied to new sensor data to
minimize effects of high non-physiological data. In the acceptance
filter, each sampled data point (e.g., every 30 seconds) is tested
for acceptance into the maximum average calculation. Each new point
is compared against the most representative estimate of the sensor
curve at the previous sampling interface (e.g., 30-second time
point), or at a projection to a current estimated value. To reject
high data, the current data point is compared to the most recent
value of the average maximum values over a time period (e.g., 5
sampled data points over a 2.5 minute period). If the ratio of
current value to the comparison value is greater than a certain
threshold (e.g., about 1.02), then the current data point is
replaced with a previously accepted value (e.g., 30-second value).
If the ratio of current value to the comparison value is in at or
within a certain range (e.g., about 1.02 to 0.90), then the current
data point is accepted. If the ratio of current value to the
comparison value is less than a certain threshold (e.g., about
0.90), then the current data point is replaced with a previously
accepted value. The acceptance filter step and max-average
calculation are continuously run throughout the data set (e.g.,
fixed 5-minute windows) on a rolling window basis (e.g., every 30
seconds).
[0356] In some implementations of the acceptance filter, the
comparison value for acceptance could also be the most recent
maximum of 5 accepted sensor points (more sensitive) or the most
recent average over 10 averages of 5 maximum values (least
sensitive), for example. In some exemplary implementations of the
acceptance filter, the projected value for the current time point
can be based on regression of the last 4 accepted 30-second values
and/or the last 2 to 4 (5 to 15 min) of the 5-minute smoothed
values, for example. In some exemplary implementations of the
acceptance filter, the percentage comparisons of +2% and -10% of
counts value would be replaced by percentage comparisons based on
the most recent 24 hour range of counts values; this method would
provide improved sensor specificity as compared to a method based
on total counts.
[0357] In another embodiment of Signal Artifacts Replacement, the
processor runs a "Cone of Possibility Replacement Method." It is
noted that this method can be performed in the sensor and/or in the
receiver. The Cone of Possibility Detection Method utilizes
physiological information along with glucose signal values in order
define a "cone" of physiologically feasible glucose signal values
within a human. Particularly, physiological information depends
upon the physiological parameters obtained from continuous studies
in the literature as well as our own observations. A first
physiological parameter uses a maximal sustained rate of change of
glucose in humans (e.g., about 4 to 5 mg/dl/min) and a maximum
sustained acceleration of that rate of change (e.g., about 0.1 to
0.2 mg/min/min). A second physiological parameter uses the
knowledge that rate of change of glucose is lowest at the maxima
and minima, which are the area of greatest risk in patient
treatment, such as described with reference to Cone of Possibility
Detection, above. A third physiological parameter uses the fact
that the best solution for the shape of the curve at any point
along the curve over a certain time period (e.g., about 20-25
minutes) is a straight line. It is noted that the maximum rate of
change can be narrowed in some instances. Therefore, additional
physiological data can be used to modify the limits imposed upon
the Cone of Possibility Replacement Method for sensor glucose
values. For example, the maximum per minute rate change can be
lower when the subject is lying down or sleeping; on the other
hand, the maximum per minute rate change can be higher when the
subject is exercising, for example.
[0358] The Cone of Possibility Replacement Method utilizes
physiological information along with blood glucose data in order to
improve the estimation of blood glucose values within a human in an
embodiment of Signal Artifacts Replacement. The Cone of Possibility
Replacement Method can be performed on raw data in the sensor, on
raw data in the receiver, or on smoothed data (e.g., data that has
been replaced/estimated in the sensor or receiver by one of the
methods described above) in the receiver.
[0359] In a first implementation of the Cone of Possibility
Replacement Method, a centerline of the cone can be projected from
a number of previous, optionally smoothed, incremental data points
(e.g., previous four, 5-minute data points). Each predicted cone
centerline point (e.g., 5 minute point) increases by the slope (S)
(e.g., for the regression in counts/minute) multiplied by the data
point increment (e.g., 5 minutes). Counts/mg/dL is estimated from
glucose and sensor range calculation over the data set.
[0360] In this first implementation of the Cone of Possibility
Replacement Method, positive and negative cone limits are simple
linear functions. Periodically (e.g., every 5 minutes), each sensor
data point (optionally smoothed) is compared to the cone limits
projected from the last four points. If the sensor value observed
is within the cone limits, the sensor value is retained and used to
generate the cone for the next data point increment (e.g., 5-minute
point). If the sensor value observed falls outside the high or low
cone limit, the value is replaced by the cone limit value, and that
value is used to project the next data point increment (e.g., 5
minute point, high point, or low point). For example, if the
difference between two adjacent 5-minute points exceeds 20 mg/dL,
then cone limits are capped at 20 mg/dL increments per 5 minutes,
with the positive limit of the cone being generated by the addition
of 0.5*A*t.sup.2 to mid cone value, where A is 0.1 mg/dL/min/min
and t is 5 minute increments (A is converted to counts/min/min for
the calculation), and the negative limit of the cone being
generated by the addition of -0.5*A*t.sup.2 to mid cone value. This
implementation provides a high degree of accuracy and is minimally
sensitive to non-physiological rapid changes.
[0361] The following Table 1 illustrates one example implementation
of the Cone of Possibility Replacement Method, wherein the maximum
sustained value observed for S is about +/-4 mg/dL/min and the
maximum value observed for A is about +/-0.1 mg/dL/min.sup.2:
TABLE-US-00001 TABLE 1 Mid line Time (mg/dL) Positive Cone Limit
Negative Cone Limit 0 100 100 100 5 100 + 5 * S 100 + 5 * S + 12.5
* A 100 + 5 * S - 12.5A 10 100 + 10 * S 100 + 10 * S + 50 * A 100 +
10 * S - 50 * A 15 100 + 15 * S 100 + 15 * S + 112.5 * A 100 + 15 *
S - 112.5 * A 20 100 + 20 * S 100 + 20 * S + 200 * A 100 + 20 * S -
200 * A 25 100 + 25 * S 100 + 25 * S + 312.5 * A 100 + 25 * S -
312.5 * A
[0362] The cone widens for each 5-minute increment for which a
sensor value fails to fall inside the cone up to 30 minutes, such
as can be seen in the table above. At 30 minutes, a cone has likely
widened enough to capture an observed sensor value, which is used,
and the cone collapses back to a 5-minute increment width. If no
sensor values are captured within 30 minutes, the cone generation
routine starts over using the next four observed points. In some
implementations special rules can be applied, for example in a case
where the change in counts in one 5-minute interval exceeds an
estimated 30-mg/dL amount. In this case, the next acceptable point
can be more than 20 to 30 minutes later. It is noted that an
implementation of this algorithm includes utilizing the cone of
possibility to predict glucose levels and alert patients to present
or upcoming dangerous blood glucose levels.
[0363] In another alternative embodiment of cone widening, the cone
can widen in set multiples (e.g., 20 mg/dL) of equivalent amounts
for each additional time interval (e.g., 5 minutes), which rapidly
widens the cone to accept data.
[0364] It is noted that the numerical parameters represent only one
example implementation of the Cone of Possibility Replacement
Method. The concepts can be applied to any numerical parameters as
desired for various glucose sensor applications.
[0365] In another implementation of the Cone of Possibility
Replacement Method, sensor calibration data is optimized using the
Clarke Error Grid, the Consensus Grid, or an alternative error
assessment that assigns risk levels based on the accuracy of
matched data pairs. In an example using the Clarke Error Grid,
because the 10 regions of the Clarke Error Grid are not all
symmetric around the Y=X perfect regression, fits to the grid can
be improved by using a multi-line regression to the data.
[0366] Accordingly the pivot point method for the counts vs.
glucose regression fit can be used to optimize sensor calibration
data to the Clarke Error Grid, Consensus Grid, or other clinical
acceptability standard. First, the glucose range is divided
according to meter values (e.g., at 200 mg/dL). Two linear fitting
lines are used, which cross at the pivot point. The coordinates of
the pivot point in counts and glucose value, plus the slope and
intercept of the two lines are variable parameters. Some of pivot
point coordinates (e.g., 4 out of 6) and slope or intercept of each
line can be reset with each iteration, while the chosen coordinates
define the remainder. The data are then re-plotted on the Clarke
Error Grid, and changes in point placement and percentages in each
region of the grid are evaluated. To optimize the fit of a data set
to a Clark Error Grid, the regression of counts vs. reference
glucose can be adjusted such that the maximum number of points are
in the A+B zones without reducing the A+B percentage, and the
number of points are optimized such that the highest percentage are
in the A zone and lowest percentage are in the D, E and C zones. In
general, the points should be distributed as evenly as possible
around the Y=X line. In some embodiments, three distinct lines
optimized for clinical acceptability can represent the regression
line. In some embodiments, an additional useful criterion can be
used to compute the root mean squared percentage bias for the data
set. Better fits are characterized by reduction in the total root
mean squared percentage bias. In an alternative implementation of
the pivot point methods, a predetermined pivot (e.g., 10 degree) of
the regression line can be applied when the estimated blood is
above or below a set threshold (e.g., 150 mg/dL), wherein the pivot
and threshold are determined from a retrospective analysis of the
performance of a conversion function and its performance at a range
of glucose concentrations.
[0367] In another embodiment of Signal Artifacts Replacement,
reference changes in electrode potential can be used to estimate
glucose sensor data during positive detection of signal artifacts
from an electrochemical glucose sensor, the method hereinafter
referred to as reference drift replacement. In this embodiment, the
electrochemical glucose sensor comprises working, counter, and
reference electrodes, such as described with reference to FIGS. 1,
2 and 10 above. This method exploits the function of the reference
electrode as it drifts to compensate for counter electrode
limitations during oxygen deficits, pH changes, and/or temperature
changes such as described in more detail above with reference to
FIGS. 10A, 10B, and 10C.
[0368] Such as described with in more detail with reference to FIG.
10A a potentiostat is generally designed so that a regulated
potential difference between the reference electrode 102 and
working electrode 100 is maintained as a constant. The potentiostat
allows the counter electrode voltage to float within a certain
voltage range (e.g., from between close to the +1.2V observed for
the working electrode to as low as battery ground or 0.0V). The
counter electrode voltage measurement will reside within this
voltage range dependent on the magnitude and sign of current being
measured at the working electrode and the electroactive species
type and concentration available in the electrolyte adjacent to the
counter electrode 104. This species will be electrochemically
recruited (e.g., reduced/accepting electrons) to equal the current
of opposite sign (e.g., oxidized/donating electrons) occurring at
the working electrode 100. It has been discovered that the
reduction of dissolved oxygen or hydrogen peroxide from oxygen
converted in the enzyme layer are the primary species reacting at
the counter electrode to provide this electronic current balance in
this embodiment. If there are inadequate reducible species (e.g.,
oxygen) available for the counter electrode, or if other
non-glucose reaction rate limiting phenomena occur (e.g.,
temperature or pH), the counter electrode can be driven in its
electrochemical search for electrons all the way to ground or 0.0V.
However, regardless of the voltage in the counter electrode, the
working and counter electrode currents must still maintain
substantially equivalent currents. Therefore, the reference
electrode 102 will drift upward creating new oxidizing and reducing
potentials that maintain equal currents at the working and counter
electrodes.
[0369] Because of the function of the reference electrode 102,
including the drift that occurs during periods of signal artifacts
(e.g., ischemia), the reference electrode can be monitored to
determine the severity of the signal artifacts on the data stream.
Particularly, a substantially direct relationship between the
reference electrode drift and signal artifacts has been discovered.
Using the information contained within the CV curve (FIGS. 10B
and/or 10C), the measured glucose signal (I.sub.SENSE) can be
automatically scaled accordingly to replace these undesired
transient effects on the data stream. It is noted that the circuit
described with reference to FIG. 10A can be used to determine the
CV curve on a regularly scheduled basis or as needed. To this end,
the desired reference voltage and applied potential are made
variable, and the reference voltage can be changed at a defined
rate while measuring the signal strength from the working
electrode, which allows for generation of a CV curve while a sensor
is in vivo.
[0370] In alternative implementations of the reference drift
replacement method, a variety of algorithms can therefore be
implemented that replaces the signal artifacts based on the changes
measured in the reference electrode. Linear algorithms, and the
like, are suitable for interpreting the direct relationship between
reference electrode drift and the non-glucose rate limiting signal
noise such that appropriate conversion to signal noise compensation
can be derived.
[0371] In other embodiments of Signal Artifacts Replacement,
prediction algorithms, also referred to as projection algorithms,
can be used to replace glucose data signals for data which does not
exist because 1) it has been discarded, 2) it is missing due to
signal transmission errors and the like, or 3) it represents a time
period (e.g., future) for which a data stream has not yet been
obtained based on historic and/or present data.
Prediction/projection algorithms include any of the above described
Signal Artifacts Replacement algorithms, and differ only in the
fact that they are implemented to replace time points/periods
during which no data is available (e.g., for the above-described
reasons), rather than including that existing data, within the
algorithmic computation.
[0372] In some embodiments, signal replacement/estimation
algorithms are used to predict where the glucose signal should be,
and if the actual data stream varies beyond a certain threshold of
that projected value, then signal artifacts are detected. In
alternative embodiments, other data processing can be applied
alone, or in combination with the above-described methods, to
replace data signals during system noise and/or signal
artifacts.
[0373] Selective Application of Signal Replacement Algorithms
[0374] FIG. 15 is a flow chart that illustrates a process of
selectively applying signal estimation in embodiments.
[0375] At block 152, a sensor data receiving module, also referred
to as the sensor data module, receives sensor data (e.g., a data
stream), including one or more time-spaced sensor data points, such
as described in more detail with reference to block 82 in FIG.
8.
[0376] At block 154, a signal artifacts detection module, also
referred to as the signal artifacts detector 154, is programmed to
detect transient non-glucose related signal artifacts in the data
stream that have a higher amplitude than system noise, such as
described in more detail with reference to block 84 in FIG. 8.
However, the signal artifacts detector of this embodiment can
additionally detect a severity of signal artifacts. In some
embodiments, the signal artifacts detector has one or more
predetermined thresholds for the severity of the signal artifacts
(e.g., low, medium, and high). In some embodiments, the signal
artifacts detector numerically represents the severity of signal
artifacts based on a calculation for example, which representation
can be used to apply to the signal estimation algorithm factors,
such as described in more detail with reference to block 156.
[0377] In one exemplary embodiment, the signal artifacts detection
module evaluates the amplitude and/or frequency of the transient
non-glucose related signal artifacts, which amplitude and/or
frequency can be used to define the severity in terms of a
threshold (e.g., high or low) or a numeric representation (e.g., a
value from 1 to 10). In another exemplary embodiment, the signal
artifacts detection module evaluates a duration of the transient
non-glucose related signal artifacts, such that as the duration
increases, a severity can be defined in terms of a threshold (e.g.,
short or long) or a numeric representation (e.g., 10, 20, 30, 40,
50, or 60 minutes). In another exemplary embodiment, the signal
artifacts detection module evaluates the frequency content from a
Fourier Transform and defines severity in terms of a threshold
(e.g., above or below 30 cycles per hour) or a numeric
representation (e.g., 50 cycles per hour). All of the signal
artifacts detection methods described herein can be implemented to
include determining a severity of the signal artifacts, threshold,
and/or numerical representations.
[0378] At block 156, the signal artifacts replacement module, also
referred to as the signal estimation module, selectively applies
one of a plurality of signal estimation algorithm factors in
response to the severity of said signal artifacts.
[0379] In one embodiment, signal artifacts replacement is normally
turned off, except during detected signal artifacts. In another
embodiment, a first signal estimation algorithm (e.g., linear
regression, FIR filter etc.) is turned on all the time, and a
second signal estimation algorithm optimized for signal artifacts
(e.g., IIR filter, Cone of Possibility Replacement Method, etc.) is
turned on only during positive detection of signal artifacts.
[0380] In another embodiment, the signal replacement module
comprises programming to selectively switch on and off a plurality
of distinct signal estimation algorithms based on the severity of
the detected signal artifacts. For example, the severity of the
signal artifacts can be defined as high and low. In such an
example, a first filter (e.g., trimmed regression, linear
regression, FIR, Reference Electrode Method, etc.) can be applied
during low signal artifacts and a second filter (e.g., IIR, Cone of
Possibility Method, etc.) can be applied during high signal
artifacts. It is noted that all of the above signal replacement
algorithms can be selectively applied in this manner based on the
severity of the detected signal artifacts.
[0381] FIG. 16 is a graph that illustrates an embodiment wherein
the signal replacement module comprises programming to selectively
switch on and off a signal artifacts replacement algorithm
responsive to detection of signal artifacts. The x-axis represents
time in minutes; the first y-axis 160 represents sensor data output
in counts. A raw data signal 161, which is illustrated as a dotted
line, shows a data stream wherein some system noise can be
detected; however signal artifacts 162 can be particularly seen in
a portion thereof. The second y-axis 164 represents
counter-electrode voltage in counts; counter electrode voltage data
165 is illustrated as a solid line. It is noted that a counter
voltage drop to approximately zero in this example, which is one of
numerous methods provided for detecting signal artifacts, detects
signal artifacts 162. Accordingly, when the system detects the
signal artifacts 162, an IIR-filter is selectively switched on in
order to replace the signal artifact with an IIR-estimated glucose
signal 166, which is illustrated as a heavy solid line. The IIR
filter is switched off upon detection of negligible signal
artifacts (e.g., counter electrode voltage increasing from about
zero in this embodiment).
[0382] FIG. 17 is a graph that illustrates an embodiment wherein
the signal artifacts replacement module comprises programming to
selectively apply different signal artifacts replacement algorithms
responsive to detection of signal artifacts. The x-axis represents
time in minutes; the first y-axis 170 represents sensor data output
in counts. A raw data signal 171, which is illustrated as a dotted
line, shows a data stream wherein some system noise can be
detected; however signal artifacts 172 can be particularly seen in
a portion thereof. The second y-axis 174 represents
counter-electrode voltage in counts; counter electrode voltage data
175 is illustrated as a solid line. It is noted that a counter
voltage drop to approximately zero in this example, which is one of
numerous methods provided for detecting signal artifacts, detects
signal artifacts 172.
[0383] In this embodiment, an FIR filter is applied to the data
stream during detection of negligible or no signal artifacts (e.g.,
during no noise to system noise in the data stream). Accordingly,
normal signal noise (e.g., system noise) can be filtered to replace
the data stream with an FIR-filtered data signal 176, which is
illustrated by a slightly heavy solid line. However, upon positive
detection of signal artifacts (e.g., detected by approximately zero
counter electrode voltage in this embodiment), the FIR filter is
switched off and an IIR-filter is switched on in order to replace
the signal artifacts with an IIR-filtered glucose signal 178, which
is illustrated as a heavy solid line. The IIR filter is
subsequently switched off and the FIR filter is switched back on
upon detection of negligible signal artifacts (e.g., counter
electrode voltage increasing from about zero in this
embodiment).
[0384] In another embodiment, the signal replacement module
comprises programming to selectively apply different parameters to
a single signal artifacts replacement algorithm (e.g., IIR, Cone of
Possibility Replacement Method, etc.). As an example, the
parameters of an algorithm can be switched according to signal
artifacts detection; in such an example, an IIR filter with a
30-minute cycle length can be used during times of no noise or
system noise and a 60-minute cycle length can be used during signal
artifacts. As another example, the severity of the signal artifacts
can be defined as short and long; in such an example, an IIR filter
with a 30-minute cycle length can be used during the short signal
artifacts and a 60-minute cycle length can be used during long
signal artifacts. As yet another example, the severity of the
signal artifacts can be defined by a numerical representation; in
such an example, the numerical representation can be used to
calculate the parameters of the signal replacement algorithm (e.g.,
IIR, Cone of Possibility Replacement Method, and Reference Drift
Method).
[0385] FIG. 18 is a flow chart that illustrates dynamic and
intelligent estimation algorithm selection process 296 in an
alternative embodiment.
[0386] At block 298, the dynamic and intelligent estimation
algorithm selection process 296 obtains sensor data, which can be
raw, smoothed, and/or otherwise processed. In some embodiments,
data matching can use data from a raw data stream received from an
analyte sensor, such as described at block 53. In some embodiments,
data matching can use calibrated data.
[0387] At block 300, the dynamic and intelligent estimation
algorithm selection process 296 includes selecting one or more
algorithms from a plurality of algorithms that best fits the
measured analyte values. In some embodiments, the estimative
algorithm can be selected based on physiological parameters; for
example, in an embodiment wherein the analyte sensor is a glucose
sensor, a first order regression can be selected when the rate of
change of the glucose concentration is high, indicating correlation
with a straight line, while a second order regression can be
selected when the rate of change of the glucose concentration is
low, indicating correlation with a curved line. In some
embodiments, a first order regression can be selected when the
reference glucose data is within a certain threshold (for example,
100 to 200 mg/dL), indicating correlation with a straight line,
while a second order regression can be selected when the reference
glucose data is outside of a certain threshold (for example, 100 to
200 mg/dL), indicating correlation with a curved line because the
likelihood of the glucose concentration turning around (for
example, having a curvature) is greatest at high and low
values.
[0388] Generally, algorithms that estimate analyte values from
measured analyte values include any algorithm that fits the
measured analyte values to a pattern, and/or extrapolates estimated
values for another time period (for example, for a future time
period or for a time period during which data needs to be
replaced). In some embodiments, a polynomial regression (for
example, first order, second order, third order, etc.) can be used
to fit measured analyte values to a pattern, and then extrapolated.
In some embodiments, autoregressive algorithms (for example, IIR
filter) can be used to fit measured analyte values to a pattern,
and then extrapolated. In some embodiments, measured analyte values
can be filtered by frequency before projection (for example, by
converting the analyte values with a Fourier transform, filtering
out high frequency noise, and converting the frequency data back to
time values by using an inverse Fourier transform); this data can
then be projected forward (extrapolated) along lower frequencies.
In some embodiments, measured analyte values can be represented
with a Wavelet transform (for example filtering out specific noise
depending on wavelet function), and then extrapolate forward. In
some alternative embodiments, computational intelligence (for
example, neural network-based mapping, fuzzy logic based pattern
matching, genetic-algorithms based pattern matching, and the like)
can be used to fit measured analyte values to a pattern, and/or
extrapolate forward. In yet other alternative embodiments,
time-series forecasting is employed using methods such as moving
average (single or double), exponential smoothing (single, double,
or triple), time series decomposition, growth curves, Box-Jenkins,
and the like. The plurality of algorithms of the preferred
embodiments can utilize any one or more of the above-described
algorithms, or equivalents, in order to intelligently select
estimative algorithms and thereby estimate analyte values.
[0389] In some embodiments, estimative algorithms further include
parameters that consider external influences, such as insulin
therapy, carbohydrate consumption, and the like. In one such
example, these additional parameters can be user input via the user
interface 47 or transmitted from an external device, such as an
insulin pump, remote device, or other computer system. By including
such external influences in additional to historical trend data
(measured analyte values), analyte concentration changes can be
better anticipated.
[0390] At block 302, the selected one or more algorithms are
evaluated based on statistical, clinical, or physiological
parameters. In some embodiments, running each algorithm on the data
stream tests each of the one or more algorithms, and the
algorithmic result with the best correlation to the measured
analyte values is selected. In some embodiments, the pluralities of
algorithms are each compared for best correlation with
physiological parameters (for example, within known or expected
rates of change, acceleration, concentration, etc). In some
embodiments, the pluralities of algorithms are each compared for
best fit within a clinical error grid (for example, within "A"
region of Clarke Error Grid). Although first and second order
algorithms are exemplified herein, any two or more algorithms such
as described in more detail below could be programmed and
selectively used based on a variety of conditions, including
physiological, clinical, and/or statistical parameters.
[0391] At block 304, the algorithm(s) selected from the evaluation
step is employed to estimate analyte values for a time period.
Accordingly, analyte values are more dynamically and intelligently
estimated to accommodate the dynamic nature of physiological data.
Additional processes, for example applying physiological
boundaries, evaluation of the estimation algorithms after employing
the algorithms, evaluating a variation of estimated analyte values,
measuring and comparing analyte values, and the like (e.g., such as
described in co-pending U.S. Published Patent Application
2005-0203360 to Brauker et al.) can be applied to the dynamic and
intelligent estimative algorithms described herein
[0392] FIG. 19 is a graph that illustrates dynamic and intelligent
estimation algorithm selection applied to a data stream in one
embodiment showing first order estimation, second order estimation,
and the measured glucose values for the time period, wherein the
second order estimation shows a better correlation to the measured
glucose data than the first order estimation. The x-axis represents
time in minutes. The y-axis represents glucose concentration in
mg/dL.
[0393] In the data of FIG. 19, measured (calibrated) sensor glucose
data 306 was obtained up to time t=0. At t=0, a first order
regression 308 was performed on the measured data 306 to estimate
the upcoming 15-minute time period. A second order regression 310
was also performed on the data to estimate the upcoming 15-minute
time period. The intelligent estimation of the preferred
embodiments, such as described in more detail below, chose the
second order regression 310 as the preferred algorithm for
estimation based on programmed conditions (at t=0). The graph of
FIG. 19 further shows the measured glucose values 312 from t=0 to
t=15 to illustrate that second order regression 310 does in fact
more accurately correlate with the measured glucose data 312 than
first order regression 308 from t=0 to t=15.
[0394] In the example of FIG. 19, the dynamic and intelligent
estimation algorithm selection determined that the second order
regression 310 was the preferred algorithm for estimation at t=0
based on conditions. A first condition was based on a set threshold
that considers second order regression a better fit when measured
glucose values are above 200 mg/dL and trending upwardly. A second
condition verifies that the curvature of the second order
regression line appropriately shows a deceleration above 200 mg/dL.
Although two specific examples of conditions are described herein,
dynamic and intelligent estimation can have as many or as few
conditions programmed therein as can be imagined or contrived. Some
additional examples of conditions for selecting from a plurality of
algorithms are listed above, however the scope of this aspect of
dynamic and intelligent estimation includes any conditional
statements that can be programmed and applied to any algorithms
that can be implemented for estimation.
[0395] FIG. 20 is a flow chart that illustrates the process 330 of
dynamic and intelligent estimation and evaluation of analyte values
in one embodiment, wherein the estimation algorithms are
continuously, periodically, or intermittently evaluated based on
statistical, clinical, or physiological parameters to maintain
accuracy of estimation.
[0396] At block 332, the dynamic and intelligent estimation and
evaluation process 130 obtains sensor data, which can be raw,
smoothed, calibrated and/or otherwise processed.
[0397] At block 334, the dynamic and intelligent estimation and
evaluation process 330 estimates one or more analyte values using
one or more estimation algorithms. In some embodiments, this
analyte value estimation uses conventional projection using first
or second order regression, for example. In some embodiments,
dynamically and intelligently selecting of one or more algorithms
from a plurality of algorithms, dynamically and intelligently
estimating analyte values within physiological boundaries,
evaluating a variation of estimated analyte values, measuring and
comparing analyte values, and the like (e.g., such as described in
U.S. Publication No. US-2005-0203360-A1) can be applied to the
dynamic and intelligent estimation and evaluation process described
herein.
[0398] The estimative algorithms described elsewhere herein
consider mathematical equations, for example, which may or may not
be sufficient to accurately estimate analyte values in some
circumstances due to the dynamic nature of mammalian behavior. For
example, in a circumstance where a patient's glucose concentration
is trending upwardly at a constant rate of change (for example, 120
mg/dL at 2 mg/dL/min), an expected physiological pattern would
likely estimate a continued increase at substantially the same rate
of change over the upcoming approximately 40 minutes, which would
fall within physiological boundaries. However, if a person with
diabetes were to engage in heavy aerobic exercise, which may not be
known by the estimative algorithm, a slowing of the upward trend,
and possibly a change to a downward trend can possibly result,
leading to inaccuracies in the estimated analyte values. Numerous
such circumstances can occur in the lifestyle of a person with
diabetes. However, although analyte values can sometimes be
estimated under "normal" circumstances, other circumstances exist
that are not "normal" or "expected" and can result in estimative
algorithms that produce apparently erroneous results, for example,
if they are based solely on mathematical calculations and/or
physiological patterns. Accordingly, evaluation of the estimative
algorithms can be performed to ensure the accuracy or quantify a
measure of confidence in the estimative algorithms.
[0399] At block 336, the dynamic and intelligent estimation and
evaluation process 330 evaluates the estimation algorithms employed
at block 334 to evaluate a "goodness" of the estimated analyte
values. The evaluation process performs an evaluation of the
measured analyte data with the corresponding estimated analyte data
(e.g., by performing the algorithm on the data stream and comparing
the measured with the corresponding analyte data for a time
period). In some embodiments, evaluation can be performed
continually or continuously so that the dynamic and intelligent
algorithms are continuously adapting to the changing physiological
analyte data. In some embodiments, the evaluation can be performed
periodically so that the dynamic and intelligent algorithms are
periodically and systematically adapting to the changing
physiological analyte data. In some embodiments, evaluation can be
performed intermittently, for example when an estimative algorithm
is initiated or when other such triggers occur, so that the dynamic
and intelligent algorithms can be evaluated when new or updated
data or algorithms are being processed.
[0400] This evaluation process 330 uses any known evaluation
method, for example based on statistical, clinical, or
physiological standards. One example of statistical evaluation is
provided below with reference to FIG. 21; however other methods are
also possible. In some embodiments, the evaluation process 330
determines a correlation coefficient of regression. In some
embodiments wherein the sensor is a glucose sensor, the evaluation
process 330 determines if the selected estimative algorithm shows
that analyte values fall with the "A" and "B" regions of the Clarke
Error Grid. Other parameters or methods for evaluation are
considered within the scope of the preferred embodiments. In some
embodiments, the evaluation process 330 includes performing a
curvature formula to determine fiducial information about the
curvature, which results in an evaluation of the amount of noise on
the signal.
[0401] In some embodiments, the evaluation process 330 calculates
physiological boundaries to evaluate whether the estimated analyte
values fall within known physiological constraints. In this
embodiment, the estimative algorithm(s) are evaluated to ensure
that they do not allow estimated analyte values to fall outside of
physiological boundaries, some examples of which are described in
more detail elsewhere herein, and in co-pending U.S. Published
Patent Application 2005-0203360 to Brauker et al., for example. In
some alternative embodiments, clinical or statistical parameters
can be used in a similar manner to bound estimated analyte
values.
[0402] If the result of the evaluation is satisfactory (for
example, 10% average deviation, correlation coefficient above 0.79,
all estimated analyte values within A or B region of the Clarke
Error Grid, all estimated analyte values within physiological
boundaries, and the like), the processing continues to the next
step, using the selected estimative algorithm. However, if the
result of the evaluation is unsatisfactory, the process can start
the algorithm selection process again, optionally considering
additional information, or the processor can determine that
estimation is not appropriate for a certain time period. In one
alternative embodiment, a signal noise measurement can be
evaluated, and if the signal to noise ratio is unacceptable, the
processor can modify its estimative algorithm or other action that
can help compensate for signal noise (e.g., signal artifacts, such
as described elsewhere herein).
[0403] FIG. 21 is a graph that illustrates an evaluation of the
selected estimative algorithm in one embodiment, wherein a
correlation is measured to determine a deviation of the measured
glucose data with the selected estimative algorithm, if any. The
x-axis represents time in minutes. The y-axis represents glucose
concentration in mg/dL. Measured glucose values 340 are shown for
about 90 minutes up to t=0. At t=0, the selected algorithm is
performed on 40 minutes of the measured glucose values 340 up to
t=0, which is represented by a regression line 342 in this
embodiment. A data association function is used to determine a
goodness of fit of the estimative algorithm on the measured glucose
data 340; namely, the estimative algorithm is performed
retrospectively on the measured glucose data 340, and is
hereinafter referred to as retrospectively estimated glucose data
342 (e.g., estimation prior to t=0), after which a correlation (or
deviation) with the measured glucose data is determined. In this
example, the goodness of fit shows a mean absolute relative
difference (MARD) of 3.3% between the measured glucose data 340 and
the retrospectively estimated glucose data 342. While not wishing
to be bound to theory, it is believed that this correlation of the
measured glucose data 340 to the retrospectively estimated glucose
data 342 can be indicative of the correlation of future estimated
glucose data to the measured glucose data for that estimated time
period.
[0404] Reference is now made to FIG. 22, which is a flow chart that
illustrates the process 450 of analyzing a variation of estimated
future analyte value possibilities in one embodiment. This
embodiment takes into consideration that analyte values are subject
to a variety of external influences, which can cause the measured
analyte values to alter from the estimated analyte values as the
time period that was estimated passes. External influences include,
but are not limited to, exercise, sickness, consumption of food and
alcohol, injections of insulin, other medications, and the like.
For a person with diabetes, for example, even when estimation does
not accurately predict the upcoming measured analyte values, the
estimated analyte values can be valuable to a patient in treatment
and in fact can even alter the estimated path by encouraging
proactive patient behavior that can cause the patient to avoid the
estimated clinical risk. In other words, the deviation of measured
analyte values from their corresponding estimated analyte values
may not be an "error" in the estimative algorithm, and is in fact
one of the benefits of the continuous analyte sensor of the
preferred embodiments, namely encouraging patient behavior
modification and thereby improving patient health through
minimizing clinically risky analyte values. Proactive behavior
modification (for example, therapies such as insulin injections,
carbohydrate consumption, exercise, and the like) can cause the
patient's measured glucose to change from the estimated path, and
analyzing a variation that can be associated with the estimated
analyte values can encompass many of these changes. Therefore, in
addition to estimated analyte values, a variation can be calculated
or estimated based on statistical, clinical, and/or physiological
parameters that provides a range of values in which the estimated
analyte values can fall.
[0405] At block 452, the variation of possible estimated analyte
values analysis process 450 obtains sensor data, which can be raw,
smoothed, calibrated and/or otherwise processed.
[0406] At block 454, the variation of possible estimated analyte
values analysis process 450 estimates one or more analyte values
using one or more estimation algorithms. In some embodiments, this
analyte values estimation uses conventional projection using first
or second order regression, for example. In some embodiments,
dynamically and intelligently selecting of one or more algorithms
from a plurality of algorithms, dynamically and intelligently
estimating analyte values within physiological boundaries, dynamic
and intelligent estimation and evaluation of estimated analyte
values, measuring and comparing analyte values (e.g., such as
described in U.S. Publication No. US-2005-0203360-A1), and the like
can be applied to the dynamic and intelligent estimation and
evaluation process described herein.
[0407] At block 456, the variation of possible estimated analyte
values evaluation process 450 analyzes a variation of the estimated
analyte data. Particularly, a statistical, clinical, and/or
physiological variation of estimated analyte values can be
calculated when applying the estimative algorithms and/or can be
calculated at regular intervals to dynamically change as the
measured analyte values are obtained. In general, analysis of
trends and their variation allows the estimation of the preferred
embodiments to dynamically and intelligently anticipate upcoming
conditions, by considering internal and external influences that
can affect analyte concentration.
[0408] In some embodiments, physiological boundaries for analytes
in mammals can be used to set the boundaries of variation. For
example, known physiological boundaries of glucose in humans are
discussed in detail with reference to U.S. Publication No.
US-2005-0203360-A1, however any physiological parameters for any
measured analyte can be implemented here to provide this variation
of physiologically feasible analyte values.
[0409] In some embodiments, statistical variation can be used to
determine a variation of possible analyte values. Statistical
variation can be any known divergence or change from a point, line,
or set of data based on statistical information. Statistical
information includes patterns or data analysis resulting from
experiments, published or unpublished, for example. In some
embodiments, statistical information can include normal patterns
that have been measured statistically in studies of analyte
concentrations in mammals, for example. In some embodiments,
statistical information can include errors observed and measured
statistically in studies of analyte concentrations in mammals, for
example. In some embodiments, statistical information can include
predetermined statistical standards, for example, deviation less
than or equal to 5% on the analyte value. In some embodiments,
statistical variation can be a measured or otherwise known signal
noise level.
[0410] In some embodiments, a variation is determined based on the
fact that the conventional blood glucose meters are known to have
up to a +/-20% error in glucose values (namely, on average in the
hands of a patient). For example, gross errors in glucose readings
are known to occur due to patient error in self-administration of
the blood glucose test. In one such example, if the user has traces
of sugar on his/her finger while obtaining a blood sample for a
glucose concentration test, then the measured glucose value will
likely be much higher than the measured glucose value in the blood.
Additionally, it is known that self-monitored blood glucose tests
(for example, test strips) are occasionally subject to
manufacturing error. In view of this statistical information, in an
embodiment wherein a continuous glucose sensor relies upon a
conventional blood glucose meter for calibration, this +/-20% error
should be considered because of the potential for translated effect
on the calibrated sensor analyte data. Accordingly, this exemplary
embodiment would provide for a +/-20% variation of estimated
glucose values based on the above-described statistical
information.
[0411] In some embodiments, a variation of estimated analyte values
can be analyzed based on individual physiological patterns.
Physiological patterns are affected by a combination of at least
biological mechanisms, physiological boundaries, and external
influences such as exercise, sickness, consumption of food and
alcohol, injections of insulin, other medications, and the like.
Advantageously, pattern recognition can be used with continuous
analyte sensors to characterize an individual's physiology; for
example the metabolism of a person with diabetes can be
individually characterized, which has been difficult to quantify
with conventional glucose sensing mechanisms due to the unique
nature of an individual's metabolism. Additionally, this
information can be advantageously linked with external influences
(for example, patient behavior) to better understand the nature of
individual human physiology, which can be helpful in controlling
the basal rate in a person with diabetes, for example.
[0412] While not wishing to be bound to theory, it is believed that
monitoring of individual historical physiological analyte data can
be used to recognize patterns that can be used to estimate analyte
values, or ranges of values, in a mammal. For example, measured
analyte data for a patient can show certain peaks of glucose levels
during a specific time of day, "normal" AM and PM eating behaviors
(for example, that follow a pattern), weekday versus weekend
glucose patterns, individual maximum rate of change, and the like,
that can be quantified using patient-dependent pattern recognition
algorithms, for example. Pattern recognition algorithms that can be
used in this embodiment include, but are not limited to, stochastic
nonlinear time-series analysis, exponential (non-linear)
autoregressive model, process feedback nonlinear autoregressive
(PFNAR) model, neural networks, and the like.
[0413] Accordingly, statistically calculated patterns can provide
information useful in analyzing a variation of estimated analyte
values for a patient that includes consideration of the patient's
normal physiological patterns. Pattern recognition enables the
algorithmic analysis of analyte data to be customized to a user,
which is useful when analyte information is variable with each
individual user, such as has been seen in glucose in humans, for
example.
[0414] In some embodiments, a variation of estimated analyte values
is on clinical risk analysis. Estimated analyte values can have
higher clinical risk in certain ranges of analyte values, for
example analyte values that are in a clinically risky zone or
analyte values that are changing at a clinically risky rate of
change. When a measured analyte value or an estimated analyte value
shows existing or approaching clinical risk, it can be important to
analyze the variation of estimated analyte values in view of the
clinical risk to the patient. For example, in an effort to aid a
person with diabetes in avoiding clinically risky hyper- or
hypoglycemia, a variation can be weighted toward the clinically
risk zone, which can be used to emphasize the pending danger to the
patient, doctor, or care taker, for example. As another example,
the variation of measured or estimated analyte values can be based
on values that fall within the "A" and/or "B" regions of an error
grid Analysis Method.
[0415] In case of variation analysis based on clinical risk, the
estimated analyte values are weighted in view of pending clinical
risk. For example, if estimated glucose values show a trend toward
hypoglycemia at a certain rate of change, a variation of possible
trends toward hypoglycemia are weighted to show how quickly the
glucose concentration could reach 40 mg/dL, for example. As another
example, if estimated glucose values show a trend toward
hyperglycemia at a certain acceleration, a variation of possible
trends toward hyperglycemia are weighted to show how quickly the
glucose concentration could reach 200 mg/dL, for example.
[0416] In some embodiments, when a variation of the estimated
analyte values shows higher clinical risk as a possible path within
that variation analysis as compared to the estimated analyte path,
the estimated analyte values can be adjusted to show the analyte
values with the most clinical risk to a patient. While not wishing
to be bound by theory, adjusting the estimated analyte values for
the highest variation of clinical risk exploits the belief that by
showing the patient the "worst case scenario," the patient is more
likely to address the clinical risk and make timely behavioral and
therapeutic modifications and/or decisions that will slow or
reverse the approaching clinical risk.
[0417] At block 458, the variation of possible estimated analyte
values evaluation process 150 provides output based on the
variation analysis. In some embodiments, the result of this
variation analysis provides a "zone" of possible values, which can
be displayed to the user, considered in data analysis, and/or used
in evaluating of performance of the estimation, for example.
[0418] FIG. 23 is a graph that illustrates variation analysis of
estimated glucose values in one embodiment, wherein a variation of
the estimated glucose values is analyzed and determined based on
known physiological parameters. The x-axis represents time in
minutes. The y-axis represents glucose concentration in mg/dL. In
this embodiment, the known maximum rate of change and acceleration
of glucose in humans are used to provide the variation about the
estimated glucose path.
[0419] The measured glucose values 460 are shown for about 90
minutes up to t=0. At t=0, intelligent and dynamic estimation of
the preferred embodiments is performed to obtain estimated glucose
values 462. A variation of estimated glucose values is then
determined based on physiological parameters, including an upper
limit 464 and a lower limit 466 of variation defined by known
physiological parameters, including rate of change and acceleration
of glucose concentration in humans.
[0420] FIG. 24 is a graph that illustrates variation of estimated
analyte values in another embodiment, wherein the variation is
based on statistical parameters. The x-axis represents time in
minutes and the y-axis represents glucose concentration in mg/dL.
The measured glucose values 470 are shown for about 160 minutes up
to t=0. At t=0, intelligent and dynamic estimation of the preferred
embodiments is employed to obtain estimated glucose values 472. A
variation is defined by upper and lower limits 474 that were
determined using 95% confidence intervals. Bremer, T.; Gough, D. A.
"Is blood glucose predictable from previous values? A solicitation
for data." Diabetes 1999, 48, 445-451, which is incorporated by
reference herein in its entirety, teaches a method of determining a
confidence interval in one embodiment.
[0421] Although some embodiments have been described for a glucose
sensor, any measured analyte pattern, data analysis resulting from
an experiment, or otherwise known statistical information, whether
official or unofficial, published or unpublished, proven or
anecdotal, and the like, can be used to provide the statistical
variation described herein.
[0422] FIG. 25 is a flow chart that illustrates the process 480 of
estimating, measuring, and comparing analyte values in one
embodiment.
[0423] At block 482, the estimating, measuring, and comparing
analyte values process 480 obtains sensor data, which can be raw,
smoothed, calibrated and/or otherwise processed.
[0424] At block 484, the estimating, measuring, and comparing
analyte values process 480 estimates one or more analyte values for
a time period. In some embodiments, this analyte values estimation
uses conventional projection using first or second order
regression, for example. In some embodiments, dynamically and
intelligently selecting of one or more algorithms from a plurality
of algorithms, dynamically and intelligently estimating analyte
values within physiological boundaries), dynamic and intelligent
estimation and evaluation of estimated analyte values, variation
analysis (e.g., such as described in co-pending U.S. Published
Patent Application 2005-0203360 to Brauker et al.), and the like
can be applied to the process described herein.
[0425] At block 486, the estimating, measuring, and comparing
analyte values process 480 obtains sensor data for the time period
for which the estimated analyte values were calculated at block
484. In some embodiments, the measured analyte data can be raw,
smoothed, calibrated, and/or otherwise processed.
[0426] At block 488, the estimating, measuring, and comparing
analyte values process 480 compares the estimated analyte data to
the measured analyte data for that estimated time period. In
general, it can be useful to compare the estimated analyte data to
the measured analyte data for that estimated time period after
estimation of analyte values. This comparison can be performed
continuously, namely, at regular intervals as data streams are
processed into measured analyte values. Alternatively, this
comparison can be performed based on events, such as during
estimation of measured analyte values, selection of a estimative
algorithm, evaluation of estimative algorithms, variation analysis
of estimated analyte values, calibration and transformation of
sensor analyte data, and the like.
[0427] One embodiment is shown in FIG. 26, wherein MARD is used to
determine a correlation (or deviation), if any, between the
estimated and measured data sets. In other embodiments, other
methods, such as linear regression, non-linear mapping/regression,
rank (for example, non-parametric) correlation, least mean square
fit, mean absolute deviation (MAD), and the like, can be used to
compare the estimated analyte data to the measured analyte data to
determine a correlation (or deviation), if any.
[0428] In one embodiment, wherein estimation is used in outlier
detection and/or in matching data pairs for a continuous glucose
sensor (see FIGS. 5 and 6), the estimated glucose data can be
plotted against reference glucose data on a clinical error grid
(for example, Clarke Error Grid or rate grid) and then compared to
the measured glucose data for that estimated time period plotted
against the same reference analyte data on the same clinical error
grid. In alternative embodiments, other clinical error analysis
methods can be used, such as Consensus Error Grid, rate of change
calculation, consensus grid, and standard clinical acceptance
tests, for example. The deviation can be quantified by percent
deviation, or can be classified as pass/fail, for example.
[0429] In some embodiments, the results of the comparison provide a
quantitative deviation value, which can be used to provide a
statistical variation; for example, if the % deviation is
calculated as 8%, then the statistical variation such as described
with reference to FIG. 22 can be updated with a +/-8% variation. In
some alternative embodiments, the results of the comparison can be
used to turn on/off the estimative algorithms, estimative output,
and the like. In general, the comparison produces a confidence
interval (for example, +/-8% of estimated values) which can be used
in data analysis, output of data to a user, and the like.
[0430] A resulting deviation from this comparison between estimated
and corresponding measured analyte values may or may not imply
error in the estimative algorithms. While not wishing to be bound
by theory, it is believed that the deviation between estimated and
corresponding measured analyte values is due, at least in part, to
behavioral changes by a patient, who observes estimated analyte
values and determines to change the present trend of analyte values
by behavioral and/or therapeutic changes (for example, medication,
carbohydrate consumption, exercise, rest, and the like).
Accordingly, the deviation can also be used to illustrate positive
changes resulting from the educational aspect of providing
estimated analyte values to the user, for example.
[0431] FIG. 26 is a graph that illustrates comparison of estimated
analyte values in one embodiment, wherein previously estimated
analyte values are compared to time corresponding measured analyte
values to determine a correlation (or deviation), if any. The
x-axis represents time in minutes. The y-axis represents glucose
concentration in mg/dL. The measured glucose values 492 are shown
for about 105 minutes up to t=15. The estimated analyte values 494,
which were estimated at t=0 for 15 minutes, are shown superimposed
over the measured analyte values 492. Using a 3-point MARD for t=0
to t=15, the estimated analyte values 494 can be compared with the
measured analyte values 492 to determine a 0.55% average
deviation.
[0432] FIG. 27 provides a flow chart 520 that illustrates the
evaluation of reference and/or sensor data for statistical,
clinical, and/or physiological acceptability in one embodiment.
Although some acceptability tests are disclosed herein, any known
statistical, clinical, physiological standards and methodologies
can be applied to evaluate the acceptability of reference and
sensor analyte data.
[0433] One cause for discrepancies in reference and sensor data is
a sensitivity drift that can occur over time, when a sensor is
inserted into a host and cellular invasion of the sensor begins to
block transport of the analyte to the sensor, for example.
Therefore, it can be advantageous to validate the acceptability of
converted sensor data against reference analyte data, to determine
if a drift of sensitivity has occurred and whether the calibration
should be updated.
[0434] In one embodiment, the reference analyte data is evaluated
with respect to substantially time corresponding converted sensor
data to determine the acceptability of the matched pair. For
example, clinical acceptability considers a deviation between time
corresponding analyte measurements (for example, data from a
glucose sensor and data from a reference glucose monitor) and the
risk (for example, to the decision making of a person with
diabetes) associated with that deviation based on the glucose value
indicated by the sensor and/or reference data. Evaluating the
clinical acceptability of reference and sensor analyte data, and
controlling the user interface dependent thereon, can minimize
clinical risk. Preferably, the receiver evaluates clinical
acceptability each time reference data is obtained.
[0435] After initial calibration, such as is described in more
detail with reference to FIG. 5, the sensor data receiving module
receives substantially continuous sensor data (e.g., a data stream)
via a receiver and converts that data into estimated analyte
values. As used herein, the term "substantially continuous" is a
broad term and is used in its ordinary sense, without limitation,
to refer to a data stream of individual measurements taken at time
intervals (e.g., time-spaced) ranging from fractions of a second up
to, e.g., 1, 2, or 5 minutes or more. As sensor data is
continuously converted, it can be occasionally recalibrated in
response to changes in sensor sensitivity (drift), for example.
Initial calibration and re-calibration of the sensor require a
reference analyte value. Accordingly, the receiver can receive
reference analyte data at any time for appropriate processing.
[0436] At block 522, the reference data receiving module, also
referred to as the reference input module, receives reference
analyte data from a reference analyte monitor. In one embodiment,
the reference data comprises one analyte value obtained from a
reference monitor. In some alternative embodiments however, the
reference data includes a set of analyte values entered by a user
into the interface and averaged by known methods, such as are
described elsewhere herein. In some alternative embodiments, the
reference data comprises a plurality of analyte values obtained
from another continuous analyte sensor.
[0437] The reference data can be pre-screened according to
environmental and physiological issues, such as time of day, oxygen
concentration, postural effects, and patient-entered environmental
data. In one exemplary embodiment, wherein the sensor comprises an
implantable glucose sensor, an oxygen sensor within the glucose
sensor is used to determine if sufficient oxygen is being provided
to successfully complete the necessary enzyme and electrochemical
reactions for accurate glucose sensing. In another exemplary
embodiment, the patient is prompted to enter data into the user
interface, such as meal times and/or amount of exercise, which can
be used to determine likelihood of acceptable reference data. In
yet another exemplary embodiment, the reference data is matched
with time-corresponding sensor data, which is then evaluated on a
modified clinical error grid to determine its clinical
acceptability.
[0438] Some evaluation data, such as described in the paragraph
above, can be used to evaluate an optimum time for reference
analyte measurement. Correspondingly, the user interface can then
prompt the user to provide a reference data point for calibration
within a given time period. Consequently, because the receiver
proactively prompts the user during optimum calibration times, the
likelihood of error due to environmental and physiological
limitations can decrease and consistency and acceptability of the
calibration can increase.
[0439] At block 524, the evaluation module, also referred to as
acceptability module, evaluates newly received reference data. In
one embodiment, the evaluation module evaluates the clinical
acceptability of newly received reference data and time
corresponding converted sensor data (new matched data pair). In one
embodiment, a clinical acceptability evaluation module 524 matches
the reference data with a substantially time corresponding
converted sensor value, and determines the Clarke Error Grid
coordinates. In this embodiment, matched pairs that fall within the
A and B regions of the Clarke Error Grid are considered clinically
acceptable, while matched pairs that fall within the C, D, and E
regions of the Clarke Error Grid are not considered clinically
acceptable.
[0440] A variety of other known methods of evaluating clinical
acceptability can be utilized. In one alternative embodiment, the
Consensus Grid is used to evaluate the clinical acceptability of
reference and sensor data. In another alternative embodiment, a
mean absolute difference calculation can be used to evaluate the
clinical acceptability of the reference data. In another
alternative embodiment, the clinical acceptability can be evaluated
using any relevant clinical acceptability test, such as a known
grid (e.g., Clarke Error or Consensus), and additional parameters,
such as time of day and/or the increase or decreasing trend of the
analyte concentration. In another alternative embodiment, a rate of
change calculation can be used to evaluate clinical acceptability.
In yet another alternative embodiment, wherein the received
reference data is in substantially real time, the conversion
function could be used to predict an estimated glucose value at a
time corresponding to the time stamp of the reference analyte value
(this can be required due to a time lag of the sensor data such as
described elsewhere herein). Accordingly, a threshold can be set
for the predicted estimated glucose value and the reference analyte
value disparity, if any. In some alternative embodiments, the
reference data is evaluated for physiological and/or statistical
acceptability as described in more detail elsewhere herein.
[0441] At decision block 526, results of the evaluation are
assessed. If acceptability is determined, then processing continues
to block 528 to re-calculate the conversion function using the new
matched data pair in the calibration set.
[0442] At block 528, the conversion function module re-creates the
conversion function using the new matched data pair associated with
the newly received reference data. In one embodiment, the
conversion function module adds the newly received reference data
(e.g., including the matched sensor data) into the calibration set,
and recalculates the conversion function accordingly. In
alternative embodiments, the conversion function module displaces
the oldest, and/or least concordant matched data pair from the
calibration set, and recalculates the conversion function
accordingly.
[0443] At block 530, the sensor data transformation module uses the
new conversion function (from block 528) to continually (or
intermittently) convert sensor data into estimated analyte values,
also referred to as calibrated data, or converted sensor data, such
as is described in more detail above.
[0444] At block 532, an output module provides output to the user
via the user interface. The output is representative of the
estimated analyte value, which is determined by converting the
sensor data into a meaningful analyte value. User output can be in
the form of a numeric estimated analyte value, an indication of
directional trend of analyte concentration, and/or a graphical
representation of the estimated analyte data over a period of time,
for example. Other representations of the estimated analyte values
are also possible, for example audio and tactile.
[0445] If, however, acceptability is determined at decision block
526 as negative (unacceptable), then the processing progresses to
block 534 to adjust the calibration set. In one embodiment of a
calibration set adjustment, the conversion function module removes
one or more oldest matched data pair(s) and recalculates the
conversion function accordingly. In an alternative embodiment, the
conversion function module removes the least concordant matched
data pair from the calibration set, and recalculates the conversion
function accordingly.
[0446] At block 536, the conversion function module re-creates the
conversion function using the adjusted calibration set. While not
wishing to be bound by theory, it is believed that removing the
least concordant and/or oldest matched data pair(s) from the
calibration set can reduce or eliminate the effects of sensor
sensitivity drift over time, adjusting the conversion function to
better represent the current sensitivity of the sensor.
[0447] At block 524, the evaluation module re-evaluates the
acceptability of newly received reference data with time
corresponding converted sensor data that has been converted using
the new conversion function (block 536). The flow continues to
decision block 538 to assess the results of the evaluation, such as
described with reference to decision block 526, above. If
acceptability is determined, then processing continues to block 530
to convert sensor data using the new conversion function and
continuously display calibrated sensor data on the user
interface.
[0448] If, however, acceptability is determined at decision block
526 as negative, then the processing loops back to block 534 to
adjust the calibration set once again. This process can continue
until the calibration set is no longer sufficient for calibration,
for example, when the calibration set includes only one or no
matched data pairs with which to create a conversion function. In
this situation, the system can return to the initial calibration or
start-up mode, which is described in more detail with reference to
FIGS. 16 and 19, for example. Alternatively, the process can
continue until inappropriate matched data pairs have been
sufficiently purged and acceptability is positively determined.
[0449] In alternative embodiments, the acceptability is determined
by a quality evaluation, for example, calibration quality can be
evaluated by determining the statistical association of data that
forms the calibration set, which determines the confidence
associated with the conversion function used in calibration and
conversion of raw sensor data into estimated analyte values. See,
e.g., U.S. Publication No. US-2005-0027463-A1.
[0450] Alternatively, each matched data pair can be evaluated based
on clinical or statistical acceptability such as described above;
however, when a matched data pair does not pass the evaluation
criteria, the system can be configured to ask for another matched
data pair from the user. In this way, a secondary check can be used
to determine whether the error is more likely due to the reference
glucose value or to the sensor value. If the second reference
glucose value substantially correlates to the first reference
glucose value, it can be presumed that the reference glucose value
is more accurate and the sensor values are errant. Some reasons for
errancy of the sensor values include a shift in the baseline of the
signal or noise on the signal due to low oxygen, for example. In
such cases, the system can be configured to re-initiate calibration
using the secondary reference glucose value. If, however, the
reference glucose values do not substantially correlate, it can be
presumed that the sensor glucose values are more accurate and the
reference glucose values eliminated from the algorithm.
[0451] FIG. 28 is a flow chart 550 that illustrates the evaluation
of calibrated sensor data for aberrant values in one embodiment.
Although sensor data are typically accurate and reliable, it can be
advantageous to perform a self-diagnostic check of the calibrated
sensor data prior to displaying the analyte data on the user
interface.
[0452] One reason for anomalies in calibrated sensor data includes
transient events, such as local ischemia at the implant site, which
can temporarily cause erroneous readings caused by insufficient
oxygen to react with the analyte. Accordingly, the flow chart 550
illustrates one self-diagnostic check that can be used to catch
erroneous data before displaying it to the user.
[0453] At block 552, a sensor data receiving module, also referred
to as the sensor data module, receives new sensor data from the
sensor.
[0454] At block 554, the sensor data transformation module
continuously (or intermittently) converts new sensor data into
estimated analyte values, also referred to as calibrated data.
[0455] At block 556, a self-diagnostic module compares the new
calibrated sensor data with previous calibrated sensor data, for
example, the most recent calibrated sensor data value. In comparing
the new and previous sensor data, a variety of parameters can be
evaluated. In one embodiment, the rate of change and/or
acceleration (or deceleration) of change of various analytes, which
have known physiological limits within the body, and sensor data
can be evaluated accordingly. For example, a limit can be set to
determine if the new sensor data is within a physiologically
feasible range, indicated by a rate of change from the previous
data that is within known physiological (and/or statistical)
limits. Similarly, any algorithm that predicts a future value of an
analyte can be used to predict and then compare an actual value to
a time corresponding predicted value to determine if the actual
value falls within a statistically and/or clinically acceptable
range based on the predictive algorithm, for example. In certain
embodiments, identifying a disparity between predicted and measured
analyte data can be used to identify a shift in signal baseline
responsive to an evaluated difference between the predicted data
and time-corresponding measured data. In some alternative
embodiments, a shift in signal baseline and/or sensitivity can be
determined by monitoring a change in the conversion function;
namely, when a conversion function is recalculated using the
equation y=mx+b, a change in the values of m (sensitivity) or b
(baseline) above a pre-selected "normal" threshold, can be used to
trigger a fail-safe or further diagnostic evaluation.
[0456] Although the above-described self-diagnostics are generally
employed with calibrated sensor data, some alternative embodiments
are contemplated that check for aberrancy of consecutive sensor
values prior to sensor calibration, for example, on the raw data
stream and/or after filtering of the raw data stream. In certain
embodiments, an intermittent or continuous signal-to-noise
measurement can be evaluated to determine aberrancy of sensor data
responsive to a signal-to-noise ratio above a set threshold. In
certain embodiments, signal residuals (e.g., by comparing raw and
filtered data) can be intermittently or continuously analyzed for
noise above a set threshold. In certain embodiments, pattern
recognition can be used to identify noise associated with
physiological conditions, such as low oxygen or other known signal
aberrancies. Accordingly, in these embodiments, the system can be
configured, in response to aberrancies in the data stream, to
trigger signal estimation, adaptively filter the data stream
according to the aberrancy, and the like, as described in more
detail herein.
[0457] In another embodiment, reference analyte values are
processed to determine a level of confidence, wherein reference
analyte values are compared to their time-corresponding calibrated
sensor values and evaluated for clinical or statistical accuracy.
In yet another alternative embodiment, new and previous reference
analyte data are compared in place of or in addition to sensor
data. In general, there exist known patterns and limitations of
analyte values that can be used to diagnose certain anomalies in
raw or calibrated sensor and/or reference analyte data.
[0458] At decision block 558, the system determines whether the
comparison returned aberrant values. In one embodiment, the slope
(rate of change) between the new and previous sensor data is
evaluated, wherein values greater than +/-10, 15, 20, 25, or 30% or
more change and/or +/-2, 3, 4, 5, 6 or more mg/dL/min, more
preferably +/-4 mg/dL/min, rate of change are considered aberrant.
In certain embodiments, other known physiological parameters can be
used to determine aberrant values. However, a variety of
comparisons and limitations can be set.
[0459] At block 560, if the values are not found to be aberrant,
the sensor data transformation module continuously (or
intermittently) converts received new sensor data into estimated
analyte values, also referred to as calibrated data.
[0460] At block 562, if the values are found to be aberrant, the
system goes into a suspended mode, also referred to as fail-safe
mode in some embodiments, which is described in more detail below
with reference to FIG. 29. In general, suspended mode suspends
display of calibrated sensor data and/or insertion of matched data
pairs into the calibration set. Preferably, the system remains in
suspended mode until received sensor data is not found to be
aberrant. In certain embodiments, a time limit or threshold for
suspension is set, after which system and/or user interaction can
be required, for example, requesting additional reference analyte
data, replacement of the electronics unit, and/or reset.
[0461] In some alternative embodiments, in response to a positive
determination of aberrant value(s), the system can be configured to
estimate one or more glucose values for the time period during
which aberrant values exist. Signal estimation generally refers to
filtering, data smoothing, augmenting, projecting, and/or other
methods for estimating glucose values based on historical data, for
example. In one implementation of signal estimation,
physiologically feasible values are calculated based on the most
recent glucose data, and the aberrant values are replaced with the
closest physiologically feasible glucose values. See also U.S.
Publication No. US-2005-0027463-A1.
[0462] FIG. 29 provides a flow chart 580 that illustrates a
self-diagnostic of sensor data in one embodiment. Although
reference analyte values can useful for checking and calibrating
sensor data, self-diagnostic capabilities of the sensor provide for
a fail-safe for displaying sensor data with confidence and enable
minimal user interaction (for example, requiring reference analyte
values only as needed).
[0463] At block 582, a sensor data receiving module, also referred
to as the sensor data module, receives new sensor data from the
sensor.
[0464] At block 584, the sensor data transformation module
continuously (or intermittently) converts received new sensor data
into estimated analyte values, also referred to as calibrated
data.
[0465] At block 586, a self-diagnostics module, also referred to as
a fail-safe module, performs one or more calculations to determine
the accuracy, reliability, and/or clinical acceptability of the
sensor data. Some examples of the self-diagnostics module are
described above, with reference block 556. The self-diagnostics
module can be further configured to run periodically (e.g.,
intermittently or in response to a trigger), for example, on raw
data, filtered data, calibrated data, predicted data, and the
like.
[0466] In certain embodiments, the self-diagnostics module
evaluates an amount of time since sensor insertion into the host,
wherein a threshold is set for the sensor's usable life, after
which time period the sensor is considered to be unreliable. In
certain embodiments, the self-diagnostics module counts the number
of times a failure or reset is required (for example, how many
times the system is forced into suspended or start-up mode),
wherein a count threshold is set for a predetermined time period,
above which the system is considered to be unreliable. In certain
embodiments, the self-diagnostics module compares newly received
calibrated sensor data with previously calibrated sensor data for
aberrant values, such as is described in more detail elsewhere
herein. In certain embodiments, the self-diagnostics module
evaluates clinical acceptability, such as is described in more
detail with reference to FIG. 28, above. In certain embodiments,
diagnostics, such as are described in U.S. Publication No.
US-2005-0161346-A1 and U.S. Publication No. US-2005-0143635-A1, can
be incorporated into the systems of preferred embodiments for
system diagnosis, for example, for identifying interfering species
on the sensor signal and for identifying drifts in baseline and
sensitivity of the sensor signal.
[0467] At block 588, a mode determination module, which can be a
part of the sensor evaluation module 524, determines in which mode
the sensor should be set (or remain). In some embodiments, the
system is programmed with three modes: 1) start-up mode; 2) normal
mode; and 3) suspended mode. Although three modes are described
herein, the preferred embodiments are limited to the number or
types of modes with which the system can be programmed. In some
embodiments, the system is defined as "in-cal" (in calibration) in
normal mode; otherwise, the system is defined as "out-of-cal` (out
of calibration) in start-up and suspended mode. The terms as used
herein are meant to describe the functionality and are not limiting
in their definitions.
[0468] Preferably, a start-up mode is provided, wherein the
start-up mode is set when the system determines that it can no
longer remain in suspended or normal mode (for example, due to
problems detected by the self-diagnostics module, such as described
in more detail above) and/or wherein the system is notified that a
new sensor has been inserted. Upon initialization of start-up mode,
the system ensures that any old matched data pairs and/or
calibration information is purged. In start-up mode, the system
initializes the calibration set, such as described in more detail
with reference to U.S. Publication No. 2006-0036142-A1. Once the
calibration set has been initialized, sensor data is ready for
conversion and the system is set to normal mode.
[0469] Preferably, a normal mode is provided, wherein the normal
mode is set when the system is accurately and reliably converting
sensor data, for example, wherein clinical acceptability is
positively determined, aberrant values are negatively determined,
and/or the self-diagnostics modules confirms reliability of data.
In normal mode, the system continuously (or intermittently)
converts (calibrates) sensor data. Additionally, reference analyte
values received by the system are matched with sensor data points
and added to the calibration set.
[0470] In certain embodiments, the calibration set is limited to a
predetermined number of matched data pairs, after which the systems
purges old or less desirable matched data pairs when a new matched
data pair is added to the calibration set. Less desirable matched
data pairs can be determined by inclusion criteria, which include
one or more criteria that define a set of matched data pairs that
form a substantially optimal calibration set.
[0471] One inclusion criterion comprises ensuring the time stamp of
the matched data pairs (that make up the calibration set) span at
least a preselected time period (e.g., three hours). Another
inclusion criterion comprises ensuring that the time stamps of the
matched data pairs are not more than a preselected age (e.g., one
week old). Another inclusion criterion ensures that the matched
pairs of the calibration set have a substantially evenly
distributed amount of high and low raw sensor data points,
estimated sensor analyte values, and/or reference analyte values.
Another criterion comprises ensuring all raw sensor data, estimated
sensor analyte values, and/or reference analyte values are within a
predetermined range (e.g., 40 mg/dL to 400 mg/dL for glucose
values). Another criterion comprises evaluating the rate of change
of the analyte concentration (e.g., from sensor data) during the
time stamp of the matched pair(s). For example, sensor and
reference data obtained during the time when the analyte
concentration is undergoing a slow rate of change can be less
susceptible to inaccuracies caused by time lag and other
physiological and non-physiological effects. Another criterion
comprises evaluating the congruence of respective sensor and
reference data in each matched data pair; the matched pairs with
the most congruence can be chosen. Another criterion comprises
evaluating physiological changes (e.g., low oxygen due to a user's
posture, position, or motion that can cause pressure on the sensor
and effect the function of a subcutaneously implantable analyte
sensor, or other effects) to ascertain a likelihood of error in the
sensor value. Evaluation of calibration set criteria can comprise
evaluating one, some, or all of the above described inclusion
criteria. It is contemplated that additional embodiments can
comprise additional inclusion criteria not explicitly described
herein.
[0472] Unfortunately, some circumstances can exist wherein a system
in normal mode can be changed to start-up or suspended mode. In
general, the system is programmed to change to suspended mode when
a failure of clinical acceptability, aberrant value check and/or
other self-diagnostic evaluation is determined, such as described
in more detail above, and wherein the system requires further
processing to determine whether a system re-start is required
(e.g., start-up mode). In general, the system will change to
start-up mode when the system is unable to resolve itself in
suspended mode and/or when the system detects a new sensor has been
inserted (e.g., via system trigger or user input).
[0473] Preferably, a suspended mode is provided wherein the
suspended mode is set when a failure of clinical acceptability,
aberrant value check, and/or other self-diagnostic evaluation
determines unreliability of sensor data. In certain embodiments,
the system enters suspended mode when a predetermined time period
passes without receiving a reference analyte value. In suspended
mode, the calibration set is not updated with new matched data
pairs, and sensor data can optionally be converted, but not
displayed on the user interface. The system can be changed to
normal mode upon resolution of a problem (positive evaluation of
sensor reliability from the self-diagnostics module, for example).
The system can be changed to start-up mode when the system is
unable to resolve itself in suspended mode and/or when the system
detects a new sensor has been inserted (via system trigger or user
input).
[0474] The systems of preferred embodiments, including a
transcutaneous analyte sensor, mounting unit, electronics unit,
applicator, and receiver for inserting the sensor, and measuring,
processing, and displaying sensor data, provide improved
convenience and accuracy because of their designed stability within
the host's tissue with minimum invasive trauma, while providing a
discreet and reliable data processing and display, thereby
increasing overall host comfort, confidence, safety, and
convenience. Namely, the geometric configuration, sizing, and
material of the sensor of the preferred embodiments enable the
manufacture and use of an atraumatic device for continuous
measurement of analytes, in contrast to conventional continuous
glucose sensors available to persons with diabetes, for example.
Additionally, the sensor systems of preferred embodiments provide a
comfortable and reliable system for inserting a sensor and
measuring an analyte level for up to 7 days or more without
surgery. The sensor systems of the preferred embodiments are
designed for host comfort, with chemical and mechanical stability
that provides measurement accuracy. Furthermore, the mounting unit
is designed with a miniaturized and reusable electronics unit that
maintains a low profile during use. The usable life of the sensor
can be extended by incorporation of a bioactive agent into the
sensor that provides local release of an anti-inflammatory, for
example, in order to slow the subcutaneous foreign body response to
the sensor.
[0475] After the usable life of the sensor (for example, due to a
predetermined expiration, potential infection, or level of
inflammation), the host can remove the sensor and mounting from the
skin, and dispose of the sensor and mounting unit (preferably
saving the electronics unit for reuse). Another sensor system can
be inserted with the reusable electronics unit and thus provide
continuous sensor output for long periods of time.
[0476] FIG. 30 is a flow chart 600 that illustrates the process of
detecting and processing signal artifacts in some embodiments.
[0477] At block 602, a sensor data receiving module, also referred
to as the sensor data module, or processor module, receives sensor
data (e.g., a data stream), including one or more time-spaced
sensor data points. In some embodiments, the data stream is stored
in the sensor for additional processing; in some alternative
embodiments, the sensor periodically transmits the data stream to
the receiver, which can be in wired or wireless communication with
the sensor. In some embodiments, raw and/or filtered data is stored
in the sensor and/or transmitted and stored in the receiver, as
described in more detail elsewhere herein.
[0478] At block 604, a signal artifacts detection module, also
referred to as the signal artifacts detector, or signal reliability
module, is programmed to detect transient non-glucose related
signal artifacts in the data stream, In some embodiments, the
signal artifacts detector can comprise an oxygen detector, a pH
detector, a temperature detector, and/or a pressure/stress
detector, for example, the signal artifacts detector 29 in FIG. 2.
In some embodiments, the signal artifacts detector is located
within the processor 22 (FIG. 2) and utilizes existing components
of the glucose sensor to detect signal artifacts, for example by
pulsed amperometric detection, counter electrode monitoring,
reference electrode monitoring, and frequency content monitoring,
which are described elsewhere herein. In yet other embodiments, the
data can be sent from the sensor to the receiver which comprises
programming in the processor 42 (FIG. 4) that performs algorithms
to detect signal artifacts, for example such as described with
reference to "Cone of Possibility Detection" method and/or by
comparing raw data vs. filtered data, both of which are described
in more detail elsewhere herein.
[0479] In some exemplary embodiments, the processor module in
either the sensor electronics and/or the receiver electronics
evaluates an intermittent or continuous signal-to-noise measurement
to determine aberrancy of sensor data responsive to a
signal-to-noise ratio above a set threshold. In some exemplary
embodiments, signal residuals (e.g., by comparing raw and filtered
data) are intermittently or continuously analyzed for noise above a
set threshold. In some exemplary embodiments, pattern recognition
can be used to identify noise associated with physiological
conditions, such as low oxygen, or other known signal aberrancies.
Accordingly, in these embodiments, the system can be configured, in
response to aberrancies in the data stream, to trigger signal
estimation, adaptively filter the data stream according to the
aberrancy, and the like, as described in more detail elsewhere
herein.
[0480] In some embodiments, one or more signal residuals are
obtained by comparing received data with filtered data, whereby a
signal artifact can be determined. In some embodiments, a signal
artifact event is determined to have occurred if the residual is
greater than a threshold. In some exemplary embodiments, another
signal artifact event is determined to have occurred if the
residual is greater than a second threshold. In some exemplary
embodiments, a signal artifact event is determined to have occurred
if the residual is greater than a threshold for a period of time or
amount of data. In some exemplary embodiments, a signal artifact
event is determined to have occurred if a predetermined number of
signal residuals above a predetermined threshold occur within a
predetermined time period (or amount of data). In some exemplary
embodiments, an average of a plurality of residuals is evaluated
over a period of time or amount of data to determine whether a
signal artifact has occurred. The use of residuals for noise
detection can be preferred in circumstances where data gaps
(non-continuous) data exists.
[0481] In some exemplary embodiments, a differential, also referred
to as a derivative of the residual, is determined by comparing a
first residual (e.g., at a first time point) and a second residual
(e.g., at a second time point), wherein a signal artifact event is
determined to have occurred when the differential is above a
predetermined threshold. In some exemplary embodiments, a signal
artifact event is determined to have occurred if the differential
is greater than a threshold for a period of time or amount of data.
In some exemplary embodiments, an average of a plurality of
differentials is calculated over a period of time or amount of data
to determine whether a signal artifact has occurred.
[0482] Numerous embodiments for detecting signal artifacts are
described in more detail in the section entitled, "Signal Artifacts
Detection," all of which are encompassed by the signal artifacts
detection at block 604.
[0483] At block 606, the processor module is configured to process
the sensor data based at least in part on whether the signal
artifact event has occurred.
[0484] In some embodiments, the sensor data is filtered in the
receiver processor to generate filtered data if the signal artifact
event is determined to have occurred; filtering can be performed
either on the raw data, or can be performed to further filter
received filtered data, or both.
[0485] In some embodiments, signal artifacts detection and
processing is utilized in outlier detection, such as described in
more detail elsewhere herein, wherein a disagreement between time
corresponding reference data and sensor data can be analyzed, e.g.,
noise analysis data (e.g., signal artifacts detection and signal
processing) can be used to determine which value is likely more
reliable (e.g., whether the sensor data and/or reference data can
be used for processing). In some exemplary embodiments wherein the
processor module receives reference data from a reference analyte
monitor, a reliability of the received data is determined based on
signal artifacts detection (e.g., if a signal artifact event is
determined to have occurred.) In some exemplary embodiments, a
reliability of the sensor data is determined based on signal
artifacts detection (e.g., if the signal artifact event is
determined to have not occurred.) The term "reliability," as used
herein, is a broad term and is used in its ordinary sense,
including, without limitation, a level of confidence in the data
(e.g., sensor or reference data), for example, a positive or
negative reliance on the data (e.g., for calibration, display, and
the like) and/or a rating (e.g., of at least 60%, 70%, 80%, 90% or
100% confidence thereon.)
[0486] In some embodiments wherein a matching data pair is formed
by matching reference data to substantially time corresponding
sensor data (e.g., for calibration and/or outlier detection)
described in more detail elsewhere herein, matching of a data pair
can be configured to occur based on signal artifacts detection
(e.g., only if a signal artifact event is determined to have not
occurred.) In some embodiments wherein the reference data is
included in a calibration factor for use in calibration of the
glucose sensor as described in more detail elsewhere herein, the
reference data can be configured to be included based on signal
artifacts detection (e.g., only if the signal artifact event is
determined to have not occurred.) In general, results of noise
analysis (e.g., signal artifact detection and/or signal processing)
can be used to determine when to use or eliminate a matched pair
for use in calibration (e.g., calibration set).
[0487] In some embodiments, a user is prompted for a reference
glucose value based on signal artifacts detection (e.g., only if a
signal artifact event is determined to have not occurred.) While
not wishing to be bound by theory, it is believed certain more
preferable times for calibration (e.g., not during noise episodes)
can be detected and processed by prompting the user for calibration
during those times.
[0488] In some embodiments, results of noise analysis (e.g., signal
artifact detection and/or signal processing) can be used to
determine how to process the sensor data. For example, different
levels of signal processing and display (e.g., raw data, integrated
data, filtered data utilizing a first filter, filtered data
utilizing a second filter, which may be "more aggressive" than the
first filter by filtering over a larger time period, and the like.)
Accordingly, the different levels of signal processing and display
can be selectively chosen responsive to a reliability measurement,
a positive or negative determination of signal artifact, and/or
signal artifacts above first and second predetermined
thresholds.
[0489] In some embodiments, results of noise analysis (e.g., signal
artifact detection and/or signal processing) can be used to
determine when to utilize and/or display different representations
of the sensor data (e.g., raw vs. filtered data), when to turn
filters on and/or off (e.g., processing and/or display of certain
smoothing algorithms), and/or when to further process the sensor
data (e.g., filtering and/or displaying). In some embodiments, the
display of the sensor data is dependent upon the determination of
signal artifact(s). For example, when a certain predetermined
threshold of signal artifacts have been detected (e.g., noisy
sensor data), the system is configured to modify or turn off a
particular display of the sensor data (e.g., display filtered data,
display processed data, disable display of sensor data, display
range of possible data values, display indication of direction of
glucose trend data, replace sensor data with predicted/estimated
sensor data, and/or display confidence interval representative of a
level of confidence in the sensor data.) In some exemplary
embodiments, a graphical representation of filtered sensor data is
displayed if the signal artifact event is determined to have
occurred. Alternatively, when a certain predetermined threshold of
signal artifacts has not been detected (e.g., minimal,
insignificant, or no noise in the data signal), the system is
configured to modify or turn on a particular display of the sensor
data (e.g., display unfiltered (e.g., raw or integrated) data, a
single data value, an indication of direction of glucose trend
data, predicted glucose data for a future time period and/or a
confidence interval representative of a level of confidence in the
sensor data.)
[0490] In some embodiments wherein a residual (or differential) is
determined as described in more detail elsewhere herein, the
residual (or differential) is used to modify the filtered data
during signal artifact event(s). In one such exemplary embodiment,
the residual is measured and then added back into the filtered
signal. While not wishing to be bound by theory, it is believed
that some smoothing algorithms may result in some loss of dynamic
behavior representative of the glucose concentration, which
disadvantage may be reduced or eliminated by the adding of the
residual back into the filtered signal in some circumstances.
[0491] In some embodiments, the sensor data can be modified to
compensate for a time lag, for example by predicting or estimating
an actual glucose concentration for a time period considering a
time lag associated with diffusion of the glucose through the
membrane, digital signal processing, and/or algorithmically induced
time lag, for example.
[0492] FIG. 31 is a graph that illustrates a raw data stream from a
glucose sensor for approximately 24 hours with a filtered version
of the same data stream superimposed on the same graph.
Additionally, this graph illustrates a noise episode, the beginning
and end of which was detected by a noise detection algorithm of the
preferred embodiments, and during which a particular filter was
applied to the data. The x-axis represents time in minutes; the
y-axis represents the raw and filtered data values in counts. In
this example, the raw data stream was obtained in 5 minute
intervals from a transcutaneous glucose sensor such as described in
more detail above, with reference to FIGS. 1B and in U.S.
Publication No. US-2006-00201087-A1.
[0493] In section 608 of the data, which encompasses an
approximately 14 hour period up to time=2:22, the filtered data was
obtained by applying a 3-point moving average window to the raw
data. During that period, the noise detection algorithm was applied
to detect a noise episode. In this example, the algorithm included
the following: calculating a residual signal by subtracting the
filtered data from the raw data (e.g., for each 5-minute point);
calculating a differential by subtracting the residual for each
5-minute point from its previous 5-minute residual; determining if
each differential exceeds a threshold of 5000 counts (and declaring
a noisy point if so); and determining whether 6 out of 12 points in
the past 1 hour exceed that threshold (and declaring a noise
episode if so). Accordingly, a noise episode was declared at
time=2:22 and a more aggressive filter was applied as described
with reference to section 610.
[0494] In section 610 of the data, also referred to as a noise
episode, which encompasses an approximately 51/2 hour period up to
time=7:57, the filtered data was obtained by applying a 7-point
moving average window to the raw data. The 7-point moving average
window was in this example was an effective filter in smoothing out
the noise in the data signal as can be seen on the graph. During
that period, an algorithm was applied to detect when the noise
episode had ended. In this example, the algorithm included the
following: calculating a residual signal by subtracting the
filtered data (using the 3-point moving average filter described
above) from the raw data (e.g., for each 5-minute point);
calculating a differential of the residual by subtracting the
residual for each 5-minute point from its previous 5-minute
residual; determining if each differential exceeds a threshold of
5000 counts (and declaring a noisy point if so); and determining
whether less than 2 noisy points had occurred in the past hour (and
declaring the noise episode over if so). Accordingly, the noise
episode was declared as over at time-7:57 and the less aggressive
filter (e.g., 3-point moving average) was again applied with the
noise detection algorithm as described with reference to section
608, above.
[0495] In section 612 of the data, which encompasses more than 4
hours of data, the filtered data was obtained by applying a 3-point
moving average window to the raw data. During that period, the
noise detection algorithm (described above) did not detect a noise
episode. Accordingly, raw or minimally filtered data could be
displayed to the patient during this time period.
[0496] It was shown that the above-described example provided
smoother glucose information during noise episodes, by applying a
more aggressive filter to smooth out the noise. It is believed that
when displayed, the smoother data will avoid presenting potentially
misleading or inaccurate information to the user. Additionally, it
was shown in the above-described example that during non-noisy
periods (when noise episodes are not detected), raw or less
aggressively filtered data can be displayed to the user in order to
provide more accurate data with minimal or no associated
filter-induced time lag in the data. Furthermore, it is believed
that proper detection of noise episodes aids in determining proper
times for calibration, ensuring more accurate calibration than may
otherwise be possible.
[0497] In the above-described example, the criteria for the onset
& offset of noise episodes were different; for example, the
onset criteria included 6 out of 12 points in the past 1 hour
exceeding a threshold, while the offset criteria included less than
2 noisy points in the past 1 hour. In this example, these different
criteria were found to create smoother transitions in the data
between the raw and filtered data and avoided false detections of
noise episodes.
[0498] FIG. 32 is a flowchart 700 that illustrates a method for
processing data from a glucose sensor in certain embodiments. In
general, prior art systems display either real-time sensor data
(e.g., prospectively calibrated/analyzed) or historical sensor data
(e.g., retrospectively calibrated/analyzed). Regarding real-time
sensor data display, the sensor data is typically prospectively
processed (e.g., calibrated, smoothed, etc) in substantially
real-time by a predetermined algorithm, wherein the real-time
prospectively processed data are displayed periodically or
substantially continuously based on that prospective analysis.
Regarding historical sensor data display, the sensor data is
typically retrospectively processed (e.g., calibrated, smoothed,
etc) after collection of an entire sensor data set, wherein the
historical retrospectively processed data are displayed based on
the retrospective analysis.
[0499] In contrast to the prior art, the preferred embodiments
describe systems and methods for periodically or substantially
continuously post-processing (e.g., updating) the substantially
real-time graphical representation of glucose data (e.g., trend
graph representative of glucose concentration over a previous
number of minutes or hours) with processed data, wherein the data
has been processed responsive to detection of signal artifacts.
[0500] At block 702, a sensor data receiving module, also referred
to as the sensor data module, or processor module, receives sensor
data (e.g., a data stream), including one or more time-spaced
sensor data points. In some embodiments, the data stream is stored
in the sensor for additional processing; in some alternative
embodiments, the sensor periodically transmits the data stream to
the receiver, which can be in wired or wireless communication with
the sensor. In some embodiments, raw and/or filtered data is stored
in the sensor and/or transmitted and stored in the receiver, as
described in more detail elsewhere herein.
[0501] At block 704, a signal artifacts detection module, also
referred to as the signal artifacts detector, or signal reliability
module, optionally detects transient non-glucose related signal
artifacts in the data stream, such as described in more detail
above with reference to block 604.
[0502] At block 706, the processor module is configured to
optionally process the sensor data based at least in part on
whether the signal artifact event has occurred, such as described
in more detail with reference to block 606 above.
[0503] At block 708, an output module, also referred to as the
processor module, provides output to the user via the user
interface. The output is representative of the estimated glucose
value, which is determined by converting the sensor data into a
meaningful glucose value such as described in more detail elsewhere
herein. User output can be in the form of a numeric estimated
glucose value, an indication of directional trend of glucose
concentration, and/or a graphical representation of the estimated
glucose data over a period of time, for example. Other
representations of the estimated glucose values are also possible,
for example audio and tactile. In some embodiments, the output
module displays both a "real-time" glucose value (e.g., a number
representative of the most recently measure glucose concentration)
and a graphical representation of the post-processed sensor data,
which is described in more detail, below.
[0504] In one embodiment, such as shown in FIG. 3A, the estimated
glucose value is represented by a numeric value. In other exemplary
embodiments, such as shown in FIGS. 3B to 3D, the user interface
graphically represents the estimated glucose data trend over
predetermined a time period (e.g., one, three, and nine hours,
respectively). In alternative embodiments, other time periods can
be represented. In some embodiments, the measured analyte value is
represented by a numeric value. In alternative embodiments, other
time periods can be represented. In alternative embodiments,
pictures, animation, charts, graphs, ranges of values, and numeric
data can be selectively displayed.
[0505] At block 710, the processor module is configured to
periodically or substantially continuously post-process (e.g.,
update) the displayed graphical representation of the data
corresponding to the time period according to the received data.
For example, the glucose trend information (e.g., for the previous
1-, 3-, or 9-hour trend graphs shown in FIGS. 3B to 3D) can be
updated to better represent actual glucose values during signal
artifacts. In some embodiments, the processor module post-processes
segments of data (e.g., 1-, 3-, or 9-hour trend graph data) every
few seconds, minutes, hours, days, or anywhere in between, and/or
when requested by a user (e.g., in responsive to a
button-activation such as a request for display of a 3-hour trend
graph screen).
[0506] In general, post-processing includes the processing
performed by the processor module (e.g., within the hand-held
receiver unit) on "recent" sensor data (e.g., data that is
inclusive of time points within the past few minutes to few hours)
after its initial display of the sensor data and prior to what is
generally termed "retrospective analysis" in the art (e.g.,
analysis that is typically accomplished retrospectively at one
time, in contrast to intermittently, periodically, or continuously,
on an entire data set, such as for display of sensor data for
physician analysis). Post-processing can include programming
performed to recalibrate the sensor data (e.g., to better match to
reference values), fill in data gaps (e.g., data eliminated due to
noise or other problems), smooth out (filter) sensor data,
compensate for a time lag in the sensor data, and the like, which
is described in more detail, below. Preferably, the post-processed
data is displayed on a personal hand-held unit (e.g., such as on
the 1-, 3-, and 9-hour trend graphs of the receiver of FIGS. 3A to
3D) in "real time" (e.g., inclusive of recent data within the past
few minutes or hours) and can be updated (post-processed)
automatically (e.g., periodically, intermittently, or continuously)
or selectively (e.g., responsive to a request) when new or
additional information is available (e.g., new reference data, new
sensor data, etc). In some alternative embodiments, post-processing
can be triggered dependent upon the duration of a noise episode;
for example, data associated with noise events extending past about
30 minutes can be processed and/or displayed differently than data
during the initial 30 minutes of a noise episode.
[0507] In one exemplary embodiment, the processor module filters
the data stream to recalculate data for a previous time period and
periodically or substantially continuously displays a graphical
representation of the recalculated data for that time period (e.g.,
trend graph).
[0508] In another exemplary embodiment, the processor module
adjusts the data for a time lag (e.g., removes a time lag induced
by real-time filtering) from data for a previous time period and
displays a graphical representation of the time lag adjusted data
for that time period (e.g., trend graph).
[0509] In another exemplary embodiment, the processor module
algorithmically smoothes one or more sensor data points over a
moving window (e.g., including time points before and after the one
or more sensor data points) for data for a previous time period and
displays a graphical representation of the updated, smoothed data
for that time period (e.g., trend graph).
[0510] Although a few examples of post-processing are described
herein, one skilled in the art appreciates a variety of data
processing that can be applied to these systems and methods,
including any of the processing steps described in more detail
elsewhere herein.
[0511] Methods and devices that are suitable for use in conjunction
with aspects of the preferred embodiments are disclosed in U.S.
Pat. No. 4,994,167; U.S. Pat. No. 4,757,022; U.S. Pat. No.
6,001,067; U.S. Pat. No. 6,741,877; U.S. Pat. No. 6,702,857; U.S.
Pat. No. 6,558,321; U.S. Pat. No. 6,931,327; and U.S. Pat. No.
6,862,465.
[0512] Methods and devices that are suitable for use in conjunction
with aspects of the preferred embodiments are disclosed in U.S.
Publication No. US-2005-0176136-A1; U.S. Publication No.
US-2005-0251083-A1; U.S. Publication No. US-2005-0143635-A1; U.S.
Publication No. US-2005-0181012-A1; U.S. Publication No.
US-2005-0177036-A1; U.S. Publication No. US-2005-0124873-A1; U.S.
Publication No. US-2005-0051440-A1; U.S. Publication No.
US-2005-0115832-A1; U.S. Publication No. US-2005-0245799-A1; U.S.
Publication No. US-2005-0245795-A1; U.S. Publication No.
US-2005-0242479-A1; U.S. Publication No. US-2005-0182451-A1; U.S.
Publication No. US-2005-0056552-A1; U.S. Publication No.
US-2005-0192557-A1; U.S. Publication No. US-2005-0154271-A1; U.S.
Publication No. US-2004-0199059-A1; U.S. Publication No.
US-2005-0054909-A1; U.S. Publication No. US-2005-0112169-A1; U.S.
Publication No. US-2005-0051427-A1; U.S. Publication No.
US-2003-0032874-A1; U.S. Publication No. US-2005-0103625-A1; U.S.
Publication No. US-2005-0203360-A1; U.S. Publication No.
US-2005-0090607-A1; U.S. Publication No. US-2005-0187720-A1; U.S.
Publication No. US-2005-0161346-A1; U.S. Publication No.
US-2006-0015020-A1; U.S. Publication No. US-2005-0043598-A1; U.S.
Publication No. US-2003-0217966-A1; U.S. Publication No.
US-2005-0033132-A1; U.S. Publication No. US-2005-0031689-A1; U.S.
Publication No. US-2004-0045879-A1; U.S. Publication No.
US-2004-0186362-A1; U.S. Publication No. US-2005-0027463-A1; U.S.
Publication No. US-2005-0027181-A1; U.S. Publication No.
US-2005-0027180-A1; U.S. Publication No. US-2006-0020187-A1; U.S.
Publication No. US-2006-0036142-A1; U.S. Publication No.
US-2006-0020192-A1; U.S. Publication No. US-2006-0036143-A1; U.S.
Publication No. US-2006-0036140-A1; U.S. Publication No.
US-2006-0019327-A1; U.S. Publication No. US-2006-0020186-A1; U.S.
Publication No. US-2006-0020189-A1; U.S. Publication No.
US-2006-0036139-A1; U.S. Publication No. US-2006-0020191-A1; U.S.
Publication No. US-2006-0020188-A1; U.S. Publication No.
US-2006-0036141-A1; U.S. Publication No. US-2006-0020190-A1; U.S.
Publication No. US-2006-0036145-A1; U.S. Publication No.
US-2006-0036144-A1; U.S. Publication No. US-2006-0016700-A1; U.S.
Publication No. US-2006-0142651-A1; U.S. Publication No.
US-2006-0086624-A1; U.S. Publication No. US-2006-0068208-A1; U.S.
Publication No. US-2006-0040402-A1; U.S. Publication No.
US-2006-0036142-A1; U.S. Publication No. US-2006-0036141-A1; U.S.
Publication No. US-2006-0036143-A1; U.S. Publication No.
US-2006-0036140-A1; U.S. Publication No. US-2006-0036139-A1; U.S.
Publication No. US-2006-0142651-A1; U.S. Publication No.
US-2006-0036145-A1; and U.S. Publication No.
US-2006-0036144-A1.
[0513] Methods and devices that are suitable for use in conjunction
with aspects of the preferred embodiments are disclosed in U.S.
application Ser. No. 09/447,227 filed Nov. 22, 1999 and entitled
"DEVICE AND METHOD FOR DETERMINING ANALYTE LEVELS"; U.S.
application Ser. No. 11/335,879 filed Jan. 18, 2006 and entitled
"CELLULOSIC-BASED INTERFERENCE DOMAIN FOR AN ANALYTE SENSOR"; U.S.
application Ser. No. 11/334,876 filed Jan. 18, 2006 and entitled
"TRANSCUTANEOUS ANALYTE SENSOR"; U.S. application Ser. No.
11/333,837 filed Jan. 17, 2006 and entitled "LOW OXYGEN IN VIVO
ANALYTE SENSOR".
[0514] All references cited herein, including but not limited to
published and unpublished applications, patents, and literature
references, are incorporated herein by reference in their entirety
and are hereby made a part of this specification. To the extent
publications and patents or patent applications incorporated by
reference contradict the disclosure contained in the specification,
the specification is intended to supersede and/or take precedence
over any such contradictory material.
[0515] The term "comprising" as used herein is synonymous with
"including," "containing," or "characterized by," and is inclusive
or open-ended and does not exclude additional, unrecited elements
or method steps.
[0516] All numbers expressing quantities of ingredients, reaction
conditions, and so forth used in the specification are to be
understood as being modified in all instances by the term "about."
Accordingly, unless indicated to the contrary, the numerical
parameters set forth herein are approximations that may vary
depending upon the desired properties sought to be obtained. At the
very least, and not as an attempt to limit the application of the
doctrine of equivalents to the scope of any claims in any
application claiming priority to the present application, each
numerical parameter should be construed in light of the number of
significant digits and ordinary rounding approaches.
[0517] The above description discloses several methods and
materials of the present invention. This invention is susceptible
to modifications in the methods and materials, as well as
alterations in the fabrication methods and equipment. Such
modifications will become apparent to those skilled in the art from
a consideration of this disclosure or practice of the invention
disclosed herein. Consequently, it is not intended that this
invention be limited to the specific embodiments disclosed herein,
but that it cover all modifications and alternatives coming within
the true scope and spirit of the invention.
* * * * *