U.S. patent application number 13/226608 was filed with the patent office on 2011-12-29 for system for characterizing chronic physiological data.
This patent application is currently assigned to Medtronic, Inc.. Invention is credited to Douglas A. Hettrick, Shantanu Sarkar, Paul D. Ziegler.
Application Number | 20110319723 13/226608 |
Document ID | / |
Family ID | 37011236 |
Filed Date | 2011-12-29 |
United States Patent
Application |
20110319723 |
Kind Code |
A1 |
Ziegler; Paul D. ; et
al. |
December 29, 2011 |
System for Characterizing Chronic Physiological Data
Abstract
An implantable medical device (IMD) senses physiological
episodes and stores data associated with the physiological episodes
in the IMD. The data is then processed based on a pattern of
recurrence of the physiological episodes.
Inventors: |
Ziegler; Paul D.;
(Minneapolis, MN) ; Hettrick; Douglas A.; (Blaine,
MN) ; Sarkar; Shantanu; (Saint Paul, MN) |
Assignee: |
Medtronic, Inc.
|
Family ID: |
37011236 |
Appl. No.: |
13/226608 |
Filed: |
September 7, 2011 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11096151 |
Mar 31, 2005 |
8036749 |
|
|
13226608 |
|
|
|
|
Current U.S.
Class: |
600/300 ;
607/62 |
Current CPC
Class: |
A61B 5/7257 20130101;
A61B 5/0031 20130101 |
Class at
Publication: |
600/300 ;
607/62 |
International
Class: |
A61B 5/00 20060101
A61B005/00; A61N 1/36 20060101 A61N001/36 |
Claims
1. A system comprising: an implantable medical device (IMD)
operable to sense and store chronic data representative of a
physiological episode; and a processing unit operable to receive
the chronic data and to characterize the chronic data based on
temporal patterns in the chronic data.
2. The system of claim 1, wherein the processing unit is further
operable to control an output of the IMD based the characterization
of the chronic data.
3. The system of claim 2, wherein the processing unit controls an
output of the IMD by establishing a schedule for reporting the
chronic data from the IMD based on the characterization of the
chronic data.
4. The system of claim 3, wherein the chronic data is characterized
by establishing a dominant frequency of repetition of the chronic
data and the reporting interval is a multiple of a dominant
frequency of repetition.
5. The system of claim 1, wherein the processing unit is further
operable to modify settings in the IMD based on the
characterization of the chronic data.
6. The system of claim 1, wherein the processing unit characterizes
the chronic data by categorizing the chronic data in a group
according to a frequency and duration of recurrence of the
physiological episode.
7. An implantable medical device (IMD) comprising: sensor circuitry
that generates chronic data representative of a physiological
episode; memory circuitry that stores the chronic data; and
processing circuitry that characterizes the chronic data based on
temporal patterns in the chronic data.
8. The IMD of claim 7, and further comprising: means for providing
an output based the characterization of the chronic data.
9. The IMD of claim 8, wherein the means for providing an output
based on the characterization of the chronic data comprises
communication circuitry operable to report the chronic data at a
reporting interval based on the temporal patterns in the chronic
data.
10. The IMD of claim 8, wherein the processing circuitry
characterizes the chronic data by determining a dominant frequency
of repetition of the chronic data and the reporting interval is a
multiple of the frequency of repetition.
11. The IMD of claim 10, wherein the dominant frequency of
repetition is determined by performing a fast Fourier transform
(FFT) on the chronic data.
12. The IMD of claim 8, wherein the means for providing an output
based the characterization of the chronic data comprises
communication circuitry for reporting a change in the temporal
patterns in the chronic data.
13. The IMD of claim 8, wherein the means providing an output based
the characterization of the chronic data comprises therapy delivery
circuitry for administering therapy based on the characterization
of the chronic data.
14. The IMD of claim 7, wherein the processing circuitry modifies
settings in the IMD based on the characterization of the chronic
data.
15. The IMD of claim 14, wherein the processing circuitry modifies
settings in the IMD based on the characterization of the chronic
data by adjusting therapy administered by the IMD based on the
characterization of the chronic data.
16. The IMD of claim 14, wherein the processing circuitry modifies
settings in the IMD based on the characterization of the chronic
data by adjusting a resolution of chronic data stored in the IMD
based on the characterization of the chronic data.
17. The IMD of claim 7, wherein the processing circuitry
characterizes the chronic data based on temporal patterns in the
chronic data by categorizing the chronic data based on a frequency
and duration of recurrence of the physiological episodes.
Description
RELATED APPLICATIONS
[0001] This application is a divisional application of U.S.
application Ser. No. 11/096,151 filed Mar. 31, 2005, now allowed,
the contents of which are incorporated by reference in its entirety
herein.
BACKGROUND OF THE INVENTION
[0002] The present invention relates to implantable medical
devices. More particularly, the present invention relates to
processing chronic physiological data from an implantable medical
device (IMD) based on patterns of recurrence in the physiological
data.
[0003] An important aspect of modern health care is the need to
monitor the vital signs and other medical episodes and data
associated with a patient, particularly those who have an IMD to
treat an illness or medical condition. This monitoring has
traditionally been performed by having a patient visit a hospital
or clinic so that a programmer or a similar device can interrogate
the IMD to gather and display the information that the IMD has
stored.
[0004] Recent developments in monitoring technology have made it
possible for a patient to upload data from an IMD to a remote
location via a communication network such as the worldwide web,
using a telephone connection or a similar type of connection to
transmit the information from the IMD to the remote location. One
system for this type of communication is the CareLink.RTM. network
provided by Medtronic, Inc. of Minneapolis, Minn. The remote
monitoring provided by such systems allows a patient with an IMD to
reduce the number and frequency of visits to a hospital or clinic,
by periodically uploading data for review by a physician or other
medical personnel to determine whether further follow up analysis
is necessary. This capability gives patients significantly greater
freedom in their lifestyle, and has brought a higher quality of
life to many patients.
[0005] For data that is best characterized by temporal patterns or
information (e.g., atrial arrhythmia burden data), the timing of
scheduled reporting of data may fail to accurately characterize the
data if it has a period of repetition longer than the reporting
interval. In these cases, diagnostic parameters extracted from the
data may differ significantly from the true value of the
parameters. In addition, it is possible that more long-term
patterns of variability will not be observed if changes in the
temporal patterns are not monitored. For example, a change in a
patient's condition may not be observed if the clinician is unable
to properly characterize the data provided by the IMD. Furthermore,
resources in the IMD (e.g., memory) may be used inefficiently if
temporal information in the data is not taken into consideration
when processing data.
BRIEF SUMMARY OF THE INVENTION
[0006] The present invention is an implantable medical device (IMD)
that senses physiological episodes and stores data associated with
the physiological episodes in the IMD. The data is then processed
based on a pattern of recurrence of the physiological episodes.
BRIEF DESCRIPTION OF THE DRAWINGS
[0007] FIG. 1 is a schematic representation of a cardiac rhythm
management system.
[0008] FIGS. 2A and 2B are bar graphs showing atrial arrhythmia
burden data over an extended period of time for two different
patients.
[0009] FIG. 3A is a bar graph showing the results of a fast Fourier
transform (FFT) on the atrial arrhythmia burden data shown in FIG.
2A.
[0010] FIG. 3B is a bar graph showing the results of a fast Fourier
transform (FFT) on the atrial arrhythmia burden data shown in FIG.
2B.
[0011] FIG. 4 is a diagram that categorizes paroxysmal atrial
arrhythmia burden data from a group of patients into related
categories of data.
[0012] FIG. 5 shows a self-organizing map (SOM) including category
models constructed from atrial arrhythmia burden data from a large
number of patients.
DETAILED DESCRIPTION
[0013] FIG. 1 is a schematic representation of cardiac rhythm
management system 20, which includes implantable medical device
(IMD) 22, remote monitor 24, patient management network 26, and
review terminals 28 and 30. IMD 22 may be a pacemaker,
defibrillator, cardioverter, pacemaker/cardioverter/defibrillator
(PCD), heart function monitor having pacing capabilities, or other
implantable device that includes the capability of providing
therapy to a patient's heart. In addition, IMD 22 may be a device
employed for continuously monitoring a patient's heart, such as the
Reveal.RTM. Insertable Loop Recorder sold by Medtronic, Inc. of
Minneapolis, Minn. IMD 22 may include therapy delivery circuitry 32
and electrogram (EGM) sensing circuitry 34, both operatively
connected to right ventricular lead 35a, left ventricular lead 35b,
and right atrial lead 35c. Therapy delivery circuitry 32 and
sensing circuitry 34 are controlled by processing circuitry 36.
Memory circuitry 37 is provided for storing sensed data. IMD 22
communicates externally with communications circuitry 38 that
communicates information wirelessly via telemetry.
[0014] Leads 35a, 35b, and 35c are positioned to provide pacing or
defibrillation pulses and sense electrical activity at desired
locations in or on the patient's heart. It will be recognized by
those skilled in the art that electrode assemblies can be
positioned at various locations that depend upon the type of
therapy provided to the patient. Each lead 35a, 35b, and 35c can
include multiple sense/pace electrodes, as well as defibrillation
coil electrodes. EGM data is sensed by measuring voltage
differentials between any pair of EGM sensing electrodes (e.g.,
tip-to-coil, tip-to-ring, and tip-to-can EGM sensing). Remote
monitor 24 is a computer or programmer that communicates with IMD
22 by telemetry, or through other wireless means and is connected
to patient management network 26 by phone or Internet connection
39. Remote monitor 24 is typically located in the patient's home,
and can interrogate IMD 22. For instance, remote monitor 24 can
initiate testing of IMD 22 at night, while the patient is sleeping,
without any direct activation by the patient. An optional phone
connection 40 can be provided with remote monitor 24 for
communicating with a technician or clinician (e.g., via a "help
line" or similar support system).
[0015] Patient management network 26 can include an
Internet-accessible server that is connected (through a local area
network, the Internet, etc.) to computers that function as review
terminals 28 and 30. Data from IMD 22 can be transmitted to patient
management network 26 via remote monitor 24, and can be stored in a
database on network 26. Terminals 28 and 30 permit patients,
healthcare providers, and technicians to access patient data to
monitor arrhythmia data on a substantially real-time basis, for
example.
[0016] A description of right ventricular lead 35a and left
ventricular lead 35b is omitted for clarity, as an understanding of
their function is not needed for an understanding of the present
invention. Ventricular leads 35a and 35b are shown only for
purposes of illustrating their connectivity with IMD 22.
[0017] In operation, leads 35a, 35b, and 35c provide therapy to a
patient and sense activations that occur during cardiac episodes.
For example, right atrial lead 35c senses atrial activations that
may occur during episodes of atrial arrhythmia, such as atrial
tachycardia (AT) and atrial fibrillation (AF). Right atrial lead
35c is electrically coupled to EGM sensing circuitry 34. EGM
sensing circuitry 34 continually monitors for episodes of atrial
arrhythmia and produces chronic data associated with the episodes
of atrial arrhythmia. The chronic data is stored in memory
circuitry 37. The chronic data stored in IMD 22 may be related to
the arrhythmia burden (i.e., how much time out of a day is spent in
a state of arrhythmia) and to the frequency of occurrence of
arrhythmia episodes.
[0018] Periodically, IMD 22 is interrogated to report the chronic
data stored in the memory circuitry of IMD 22 for analysis. Remote
monitor 24 may be a programmer at a clinician's office or a remote
device for uploading data via a communications network, such as the
worldwide web. IMD 22 includes communication circuitry 38 that
communicates information wirelessly with remote monitor 24 via
telemetry signals.
[0019] Upon interrogation, IMD 22 transmits information to remote
monitor 24 relating to the operation of EGM sensing circuitry 34,
such as diagnostic information, sensed conditions associated with
the patient (including the chronic data relating to the arrhythmia
burden and to the frequency of occurrence of arrhythmia episodes),
or any other information collected or identified by IMD 22.
[0020] When data is best characterized by temporal patterns (e.g.,
atrial arrhythmia burden data), the timing of the scheduled
reporting may fail to accurately characterize the data if it has a
period of repetition longer than the reporting interval. To
illustrate, FIGS. 2A and 2B show bar graphs of total atrial
arrhythmia burden (that is, how much time out of a day is spent in
a state of atrial tachycardia or atrial fibrillation) obtained from
PCDs implanted in patients A and B, respectively. The atrial
arrhythmia burden in each patient was measured over an extended
period of time (approximately one year). Patient A experienced
episodes of atrial arrhythmia less frequently than Patient B (that
is, the percentage of time in atrial arrhythmia was less for
Patient A), but the overall burden of each episode was more
significant for Patient A than for Patient B. Due to the
differences in frequency of occurrence of the arrhythmia episodes
between Patient A and Patient B, an appropriate interval for
reporting data stored in IMD 22 is not the same for Patient A and
Patient B. In other words, if the same reporting interval is used
for Patient A and Patient B (e.g., weekly), the chronic data
reported by each patient may fail to accurately characterize short-
and long-term patterns of repetition and variability within the
data. As a result, the chronic data reported for a particular
reporting interval may fall between episodes of atrial
arrhythmia.
[0021] In an embodiment of the present invention, the chronic data
is adaptively reported for analysis and diagnosis based on a
pattern or cycle of recurrence of physiological episodes or events.
For example, an appropriate interval for presenting the chronic
data may be determined by establishing the dominant frequency of
repetition for the chronic data. One approach to determining the
dominant frequency of repetition for the data is by performing a
fast Fourier transform (FFT) on the data. An FFT is a simplified
form of a discrete Fourier transform, which converts time domain
data into frequency domain data. The simplified algorithm uses less
processing resources than the more complex discrete Fourier
transform. This is because the FFT requires only 2N log N
calculations, while a discrete Fourier transform requires 2N.sup.2
calculations (where N is the number of discrete time samples).
[0022] FIGS. 3A and 3B show bar graphs of FFTs performed on chronic
atrial arrhythmia data for Patient A (FIG. 2A) and Patient B (FIG.
2B), respectively. The dominant frequency (or period) of repetition
can be extracted from the graph by manipulating the frequency
component with the highest amplitude. For Patient A, the highest
amplitude occurs at frequency component 50 (which is at a frequency
of 36/day). Based on the number of samples taken, this translates
to a dominant period of repetition of 14.2 days. For Patient B, the
highest amplitude occurs at frequency component 52 (which is at a
frequency of 48/day). Based on the number of samples taken, this
translates to a dominant period of repetition of 5.3 days. Thus, if
both Patients A and B are required to provide data stored in their
respective IMDs on a weekly basis, for example, the data provided
by Patient B will more likely be representative of the underlying
temporal patterns than the data provided by Patient A.
[0023] Once the dominant frequency or period of repetition has been
determined for the chronic data, the health care provider in charge
of analyzing the data and programming IMD 22 may change the
interval for reporting the chronic data to assure that the chronic
data reported characterizes the underlying temporal patterns. In
one embodiment, the reporting interval is a multiple of the
dominant period of repetition of the chronic data. When the chronic
data is reported at intervals based on the dominant period of
repetition of the chronic data, clinical decisions regarding
programming the IMD for therapy are more likely to be in agreement
with the underlying condition being treated. In the case of the
chronic atrial arrhythmia burden data, the clinician is more likely
to know the extent of the burden in both Patient A and Patient B
(and the effects therapy has on the arrhythmia) if the interval for
reporting the chronic data is based on the dominant period of
repetition of the chronic data. As a result, the clinician can
classify the patient's burden extent and can reprogram IMD 22 as
necessary to deliver appropriate therapies.
[0024] The temporal patterns or cycles of many physiological events
or episodes (including atrial arrhythmia burden) have a tendency to
fluctuate over time as administered therapies either treat or fail
to treat the underlying physiological condition. In the latter
case, these fluctuations may be indicative of an underlying change
in the condition or progression of the disease being treated, which
may necessitate medical intervention (such as a change in therapy
delivered). In addition, other physiological conditions in the
patient may impact the temporal patterns of the physiological
condition being treated. Consequently, an appropriate reporting
interval for a particular patient will change as the physiological
condition changes. The present invention overcomes these potential
errors by accounting for these changes through adaptively varying
the interval for reporting chronic data to assure that the reported
data correctly represents the temporal patterns in the chronic
data. In particular, each time chronic data from IMD 22 is reported
to remote monitor 24, remote monitor 24 determines whether the
dominant frequency of repetition of the chronic data has changed
since the previous reporting period. If the dominant frequency of
repetition of the chronic data has changed, a new reporting
interval is established based on the new dominant frequency of
repetition (either by the health care provider or automatically by
remote monitor 24).
[0025] In an alternative embodiment, processing circuitry 36 of IMD
22 continually updates the dominant frequency of repetition of the
chronic data stored in memory circuitry 37. For example, as chronic
data is sensed and stored in IMD 22, processing circuitry 36
periodically performs an FFT on the stored data to continually
reestablish the dominant frequency of repetition of the data. IMD
22 may be programmed to adaptively transmit data (or alert the
patient when a transmission is appropriate) at intervals based on
the dynamically updated dominant frequency of repetition. In
addition, IMD 22 may be programmed to alert the patient or
clinician when a change in the dominant frequency of repetition
occurs, which may be indicative of a change in the underlying
physiological condition being treated. For example, if the change
in the dominant frequency of repetition is indicative of a
deteriorating physiological condition, IMD 22 may be programmed to
emit an audible alarm relating to the extent of the change in
condition or indicating that a different treatment strategy may
need to be employed. Likewise, if the change in the dominant
frequency of repetition is indicative of an improvement in the
underlying physiological condition, IMD 22 may be programmed to
alert the patient or clinician more passively (e.g., by emitting a
signal to remote unit 24).
[0026] Besides establishing an appropriate reporting interval for
the chronic data, the chronic data may also be characterized to
optimize the operation of IMD 22. That is, remote monitor 24 may
adjust settings or programming parameters in IMD 22, or IMD 22 may
do so automatically, based on the characterized chronic data. In
one embodiment, use of memory circuitry 37 in IMD 22 is optimized
by adjusting the resolution of data stored by memory circuitry 37
based on recurrence patterns in the data. For example, in a patient
having infrequently occurring physiological episodes, the available
memory in memory circuitry 37 is allocated to record a large amount
of detail for each episode. In contrast, in a patient having
frequently occurring physiological episodes, IMD 22 is more
selective in allocation of memory circuitry 37 since data for more
events must be stored. In this embodiment, when IMD 22 reports
chronic data, memory circuitry 37 preferably has a minimal amount
of unused space (e.g., one byte).
[0027] In addition to assuring that chronic data obtained from IMD
22 accurately characterizes patterns or cycles of repetition in the
data, it is also important to be able to characterize the chronic
data in terms of long-term patterns based on the frequency and
duration of recurrence of the physiological condition. In another
embodiment of the present invention, chronic data from IMDs in a
large sampling of patients is categorized based on temporal
patterns in the data. A clinician may use this categorization to
determine an appropriate interval for reporting of the chronic
data, to more accurately characterize and diagnose the underlying
condition, and to administer an appropriate therapy regimen.
[0028] FIG. 4 shows another embodiment of the present invention,
showing a diagram that categorizes paroxysmal and persistent atrial
arrhythmia burden patterns from a sampling of patients using a
self-organizing map (SOM). For a thorough explanation of
self-organizing maps, see Kohonen, T. (1995), Self-Organizing Maps,
Series in Information Sciences, Vol. 30, Springer, Heidelberg, 2nd
Ed. (1997). In short, an SOM constructs category models from a
large collection of data, such as chronic data from a large number
of patients. To illustrate, FIG. 5 shows an example SOM including a
nine-by-nine arrangement of category models constructed from atrial
arrhythmia burden data from a large database of patient data. The
category models are arranged such that neighboring category models
in the SOM have similar characteristics. The SOM in FIG. 5 is used
to organize chronic data from the sampling of patients into the
categories shown in FIG. 4 based on the category models. The
criteria for each category model changes dynamically as data from
more patients is integrated into the database.
[0029] In the embodiment shown in FIG. 4, four categories of
paroxysmal atrial arrhythmia are identified (categories 60, 62, 64,
and 66), and one category of persistent atrial arrhythmia is
identified (category 68) from the sampling of patients based on the
category models in FIG. 5. Paroxysmal atrial arrhythmia is a
sporadic arrhythmia that is characterized by an abrupt onset and
termination of arrhythmia episodes. Persistent atrial arrhythmia is
an arrhythmia that is characterized by episodes of continuous
burden for several days (typically more than seven). The burden
data in FIG. 4 was collected by measuring the total burden for a
day in each patient for a period of 120 days. Each category 60-68
categorizes the atrial arrhythmia burden data based on their daily
burden patterns. This categorization is performed either manually
by a health care provider with the aid of visual representations of
the category models in the SOM or automatically by quantitative
analysis of the data relative to the SOM. Category 60 includes
patients having daily burden that occurs occasionally over time,
category 62 includes patients having small daily burden occurring
consistently almost every day, category 64 includes patients having
large daily burden occurring frequently over time, and category 66
includes patients having large daily burden occurring consistently
everyday. In another embodiment, the categories in the SOM are
determined based on the frequency of atrial arrhythmia episodes. In
a further embodiment, the categories in the SOM are determined
based on a combination of the daily burden and episode
frequency.
[0030] The diagram shown in FIG. 4 may be produced from an SOM by
organizing chronic data from a group of patients into categories
based on temporal patterns in the data, attributes of temporal
patterns (e.g., mean, coefficient of variation, dominant frequency,
etc.), or other attributes (e.g., patient demographics, therapy
regimen, etc.). When chronic data from a group of patients has been
categorized using an SOM or a neural network technique, the
categories may be presented to a clinician for analysis. In one
embodiment, the chronic data from a group of patients is stored in
patient management network 26 and is categorized for analysis and
display on review terminals 28 and 30.
[0031] While four categories are shown in FIG. 4, any number of
categories may be included in the diagram in accordance with the
present invention. In addition, while the exemplary embodiment
describes the use of an SOM to organize the chronic data, other
data processing techniques (such as supervised or unsupervised
neural networks, k-means clustering, principal component analysis,
independent component analysis, and decision trees) may also be
used to arrive at a similar categorization system.
[0032] Once an SOM of data has been established, a clinician may
correlate chronic data for individual patients with one of the
categories derived from the category models in the SOM. This may
provide the clinician with a guideline for determining an
appropriate reporting interval for the chronic data stored in IMD
22. For example, the clinician may assign a reporting interval to
each category that accurately characterizes the data in that
category. Thus, when a patient's chronic data is incorporated into
the proper category, a reporting interval that is appropriate for
that patient's data is immediately known.
[0033] In addition, the category in which the chronic data is
placed may provide the clinician with a guideline for diagnosis of
the underlying condition or for prescribing an appropriate therapy
based on the categorization. That is, the clinician may associate
certain categories with particular therapies that are effective
with patients whose data falls within that category. For example, a
clinician may determine that patients whose chronic data is
categorized in category 60 and 64 have arrhythmias that are more
likely due to focal triggers, and thus respond well to pulmonary
vein isolation. Thus, if a patient's chronic data falls into or
moves to categories 60 or 64, then pulmonary vein isolation is
likely the best therapy for that patient.
If a patient's condition changes such that a change in the
categorization of the patient's chronic data occurs, the chronic
data is adaptively categorized in the appropriate category. For
example, in the embodiment shown in FIG. 4, a patient's daily
burden may change from occurring occasionally over time to a small
daily burden occurring consistently almost every day (necessitating
a change from category 60 to 62). This may be indicative of a
change in the underlying condition being treated, and a
corresponding change in therapy may be necessary. Since different
reporting intervals are assigned to each category, this would also
result in a change in the reporting interval for that patient. In
summary, for chronic data that is best characterized by temporal
patterns, diagnostic parameters extracted from the data may differ
significantly from the true value of the parameters and long-term
patterns of variability may not be observable. The present
invention is a system and method for processing chronic data
collected by an implantable medical device (IMD) based on a pattern
of recurrence of physiological episodes. The IMD continually senses
physiological episodes and stores data associated with the
physiological episodes in the IMD. The data is then processed based
on a pattern of recurrence of the physiological episodes.
[0034] Although the present invention has been described with
reference to preferred embodiments, workers skilled in the art will
recognize that changes may be made in form and detail without
departing from the spirit and scope of the invention. For example,
while the embodiments described have been directed to the
characterization of atrial arrhythmia burden data, the present
invention may be applied to chronic data associated with the
treatment of other physiological conditions, such as congestive
heart failure. The present invention may also be applied to any
type of chronic physiological data that is characterizable based on
temporal patterns in the data (e.g., hemodynamic changes, patient
activity, heart rate at day and night, heart rate variability,
systolic, diastolic, and pulse pressure, etc.). The characterized
chronic data may then be reported based on patterns of recurrence
or repetition in the data. Also, changes in the data may be
monitored and analyzed to determine whether they are related to a
change in the underlying condition being treated. Further, the
device may be reprogrammed to alter the therapy delivered by the
medical device.
[0035] In addition, all embodiments described have been directed to
chronic data produced by IMD 22, a
pacemaker/cardioverter/defibrillator (PCD). It will be appreciated
that the present invention is not limited to the management of
chronic data produced by a PCD. The methods of the present
invention are applicable to any type of IMD including, but not
limited to, a cardiac pacemaker, a defibrillator, a muscular
stimulator, a brain stimulator, a nerve stimulator, a drug delivery
device, an implantable loop recorder, or a physiological monitor.
In essence, any device that produces chronic data that is
characterizable based on temporal patterns may be used in
accordance with the present invention.
[0036] Furthermore, other signal processing algorithms and
techniques may be employed to characterize the chronic data based
on temporal patterns. For example, the coefficient of variation of
the chronic data (i.e., the average divided by the standard
deviation of the data) may be determined to establish the
repeatability of the data. More specifically, a lower coefficient
of variation for a set of chronic data indicates that the patient's
physiological episodes occur more frequently or consistently, while
a higher coefficient of variation indicates that the patient's
physiological episodes occur less frequently or consistently.
Consequently, the coefficient of variation may be used to
adaptively determine an appropriate reporting interval for the
chronic data. In addition, changes in the coefficient of variation
may be indicative of a change in the underlying condition being
treated, which may require a change in the therapy administered by
the device.
* * * * *