U.S. patent application number 17/651554 was filed with the patent office on 2022-09-01 for wellness analysis system.
The applicant listed for this patent is Masimo Corporation. Invention is credited to Ammar Al-Ali.
Application Number | 20220277849 17/651554 |
Document ID | / |
Family ID | |
Filed Date | 2022-09-01 |
United States Patent
Application |
20220277849 |
Kind Code |
A1 |
Al-Ali; Ammar |
September 1, 2022 |
WELLNESS ANALYSIS SYSTEM
Abstract
A wellness analyzer is in communications with sensors that
generate real-time physiological data from a patient. The wellness
analyzer is also in communications with databases that provide
non-real-time information relevant to a medical-related assessment
of the patient. In a diagnostic mode, a monitor layer inputs the
sensor data and adjunct layers input the database information.
Adjunct layer logic blocks process the database information so as
to output supplemental information to the monitor. Monitor logic
blocks process the sensor data and the supplemental information so
as to generate a wellness output. In a simulation mode, a simulator
generates at least one parameter and the monitor generates a
predictive wellness output accordingly.
Inventors: |
Al-Ali; Ammar; (San Juan
Capistrano, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Masimo Corporation |
Irvine |
CA |
US |
|
|
Appl. No.: |
17/651554 |
Filed: |
February 17, 2022 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13009505 |
Jan 19, 2011 |
11289199 |
|
|
17651554 |
|
|
|
|
61296467 |
Jan 19, 2010 |
|
|
|
International
Class: |
G16H 50/20 20060101
G16H050/20; G16H 50/50 20060101 G16H050/50 |
Claims
1.-20. (canceled)
21. A wellness analyzer system comprising: a plurality of sensors
configured to generate real-time physiological data from a patient;
a plurality of databases configured to provide non-real-time
information relevant to a medical-related assessment, the plurality
of databases comprising patient-specific databases and
non-patient-specific databases; and a wellness monitor including
one or more processors, wherein the one or more processors are
configured to receive the real-time physiological data and the
non-real-time information, process the real-time physiological data
to obtain a plurality of physiological parameters, and process the
non-real-time information to generate supplemental information, the
wellness monitor further configured to: in a diagnostic mode,
generate a wellness output based at least in part on one or more
features of the plurality of physiological parameters and the
supplemental information; and store the one or more features in a
memory of the wellness monitor; and in a predictive mode, construct
a virtual patient model from at least the non-real-time information
and test the virtual patient model to assess a health risk of the
patient in response to predetermined physical, medical or
environmental conditions, wherein the wellness monitor is
configured to test the virtual patient model based on one or more
simulated physiological parameters and the one or more features
stored in the memory.
22. The system of claim 21, wherein the virtual patient model is
constructed further based on a history of patient monitoring
information.
23. The system of claim 21, wherein the one or more processors
comprise a monitor logic layer configured to extract the one or
more features of the plurality of physiological parameters.
24. The system of claim 23, wherein the one or more processors
further comprise adjunct logic layers configured to process the
non-real-time information to generate the supplemental
information.
25. The system of claim 24, wherein each of the adjunct layers is
specialized to process data according to a source of the data.
26. The system of claim 21, wherein the health risk is a risk of
infection, septic shock, embolism, pneumonia, or adverse drug
reaction.
27. The system of claim 21, wherein the health risk is a long-term
susceptibility to a disease.
28. The system of claim 21, wherein the wellness monitor is
configured to output an assessment of the health risk as a
predictive wellness index.
29. The system of claim 21, wherein the wellness monitor is
configured to output an assessment of the health risk in a listing
of specific health risks.
30. The system of claim 29, wherein the listing of specific health
risks is sorted according to priority or urgency for a follow-up
examination or treatment.
31. The system of claim 21, wherein the plurality of sensors are
configured to measure at least one of an oxygen saturation, pulse
rate, total hemoglobin, temperature, or respiration rate of the
patient.
32. The system of claim 21, wherein the plurality of sensors
comprise a plethysmographic sensor.
33. The system of claim 21, wherein the patient specific databases
comprise a medical history of the patient.
34. The system of claim 21, wherein the patient specific databases
comprise a lab report or a test result of the patient.
35. The system of claim 21, wherein the patient specific databases
comprise one or more of prescribed medications or prescribed
therapies of the patient.
36. The system of claim 21, wherein the patient specific databases
comprise a family history, genealogy, genetic information, or
environmental information of the patient.
37. The system of claim 21, wherein the non-patient-specific
databases comprise scientific research information.
38. The system of claim 21, wherein the non-real-time information
comprises local databases in communication with the wellness
monitor via a local area network.
39. The system of claim 21, wherein the non-real-time information
comprises non-local databases in communication with the wellness
monitor via a wide area network.
Description
REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of U.S. patent
application Ser. No. 13/009,505, filed Jan. 19, 2011, titled
"WELLNESS ANALYSIS SYSTEM," which claims priority benefit under 35
U.S.C. .sctn. 119(e) to U.S. Provisional Patent Application Ser.
No. 61/296,467, filed Jan. 19, 2010, titled "WELLNESS ANALYSIS
SYSTEM," which is incorporated in its entirety by reference
herein.
BACKGROUND OF THE INVENTION
[0002] Various monitoring technologies are available for assessing
one or more physiological systems. For example, pulse oximetry is a
widely accepted noninvasive procedure for measuring blood oxygen
saturation and pulse rate, which are significant indicators of
circulatory system status. Pulse oximeters capable of reading
through motion induced noise are disclosed in at least U.S. Pat.
Nos. 6,770,028, 6,658,276, 6,650,917, 6,157,850, 6,002,952,
5,769,785 and 5,758,644; low noise pulse oximetry sensors are
disclosed in at least U.S. Pat. Nos. 6,088,607 and 5,782,757; all
of which are assigned to Masimo Corporation, Irvine, Calif.
("Masimo") and are incorporated by reference herein.
[0003] Physiological monitors and corresponding multiple wavelength
optical sensors are described in at least U.S. patent application
Ser. No. 11/367,013, filed Mar. 1, 2006 and titled Multiple
Wavelength Sensor Emitters and U.S. patent application Ser. No.
11/366,208, filed Mar. 1, 2006 and titled Noninvasive Multi
Parameter Patient Monitor, both assigned to Masimo Laboratories,
Irvine, Calif. (Masimo Labs) and both incorporated by reference
herein.
[0004] Further, physiological monitoring systems that include low
noise optical sensors and pulse oximetry monitors, such as any of
LNOP.RTM. adhesive or reusable sensors, SofTouch.TM. sensors, Hi-Fi
Trauma.TM. or Blue.TM. sensors; and any of Radical.RTM.,
SatShare.TM., Rad-9.TM., Rad-5.TM., Rad-5v.TM. or PPO+.TM. Masimo
SET.RTM. pulse oximeters, are all available from Masimo.
Physiological monitoring systems including multiple wavelength
sensors and corresponding noninvasive blood parameter monitors,
such as Rainbow.TM. adhesive and reusable sensors and RAD-57.TM.
and Radical-7.TM. monitors for measuring SpO.sub.2, pulse rate
(PR), perfusion index (PI), pleth variability index (PVI), signal
quality, HbCO, HbMet and Hbt among other parameters are also
available from Masimo.
[0005] Various other monitoring technologies can assess the status
of other physiological systems. U.S. Pat. App. Publication No.
2008/0108884 A1 filed Sep. 24, 2007 and titled Modular Patient
Monitor describes monitoring of blood constituent parameters and
respiration rate as well as blood pressure, blood glucose, ECG,
CO.sub.2 and EEG to name a few and is also assigned to Masimo and
incorporated by reference herein.
SUMMARY OF THE INVENTION
[0006] Physiological monitoring is advanced with any methodology
that integrates the many otherwise disparate monitoring
technologies and measured parameters. Further, physiological
monitoring may be enhanced by taking into account other patient
data including medical history, medications, lab work, diagnostic
test results among other data outside of real-time patient
monitoring. Also, volumes of medical research and vast databases of
scientific knowledge can be accessed to further supplement and
integrate the results of real-time monitoring.
[0007] A wellness analysis system advantageously integrates
real-time sensor data regarding the status of any or all of a
body's circulatory, respiratory, neurological, gastrointestinal,
urinary, immune, musculoskeletal, endocrine and reproductive
systems so as to generate a wellness output. In addition, a
wellness analysis system stores traces of measured parameters
during real-time patient monitoring so as to characterize a
patient's response over time, creating a "virtual patient" that can
be tested with simulated data, resulting in a predictive wellness
output.
[0008] An aspect of a wellness analyzer is sensors that generate
real-time physiological data from patient sites. Further, databases
provide non-real-time information relevant to a medical-related
assessment of the patient. A wellness monitor, in a diagnostic
mode, inputs the sensor data. Adjunct layers input the database
information. Adjunct layer logic blocks process the database
information so as to output supplemental information to the
wellness monitor. Wellness monitor logic blocks process the sensor
data and supplemental information so as to generate a wellness
output.
[0009] Another aspect of a wellness analyzer is physiological
sensor data input to a monitor layer. Physiological parameters are
derived based upon the sensor data. Physiological system statuses
are generated based at least in part upon the parameters. A
wellness output is based upon these physiological system
statuses.
[0010] A further aspect of a wellness analyzer is a monitor layer
means for generating a wellness index responsive to real-time
sensor data. An adjunct layer means is coupled to the monitor layer
means for supplementing the monitor layer wellness index according
to non-real-time data.
BRIEF DESCRIPTION OF THE DRAWINGS
[0011] FIG. 1 is a general block diagram of a wellness analysis
system;
[0012] FIG. 2 is a general flow diagram of a wellness analyzer
embodiment;
[0013] FIG. 3 is a detailed flow diagram of a wellness analyzer
embodiment;
[0014] FIG. 4 is a detailed flow diagram of a wellness monitor
embodiment;
[0015] FIG. 5 is a block diagram of a parameter logic block;
[0016] FIG. 6 is a block diagram of a system logic block;
[0017] FIG. 7 is a block diagram of a diagnostic logic block;
and
[0018] FIG. 8 is a general flow diagram of a predictive wellness
monitor embodiment.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0019] FIG. 1 illustrates a wellness analysis system 100 having a
wellness analyzer 110 in communications with various forms of
patient data. In particular, sensors 112 provide real-time patient
data responsive to one or more of a patient's circulatory,
respiratory, neurological, gastrointestinal, urinary, immune,
musculoskeletal, endocrine and reproductive systems. Further,
various local, regional, nationwide or worldwide databases provide
non-real-time patient-related data. Local databases include
hospital records 120 that communicate with the wellness analyzer
110 via a local area network (LAN) 130. Hospital records 120 may
include patient history 122, recent lab work and tests 124 and
prescribed medications and therapies 128, to name a few. Non-local
databases 140 include patient-specific data 142 and non-specific
data 144 that are communicated to the wellness analyzer 110 via a
wide area network (WAN) 150. Patient-specific data 142 may include
family history, genealogy, genetic and environmental information,
as a few examples. Non-specific data 144 may include scientific
research and other information that generally relates to known
patient physiology and history, such as genetic, pharmacological,
medical and environmental research, to name a few.
[0020] Advantageously, the wellness analyzer 110, in a diagnostic
mode, utilizes the sensor data 112 in conjunction with historical
data 120 and research 140 so as to derive a comprehensive wellness
output 116 for a patient. In an embodiment, the wellness output 116
is a simple index useful for an initial screening of individuals.
For example, the index may be a range of values, such as 1 through
10, with 10 indicating no significant health issues and 1
indicating the existence of one or more potentially
life-threatening conditions. In another embodiment, the wellness
output 116 is a comprehensive and detailed listing of specific
health issues sorted according to priority or urgency for follow-up
examination, observation and treatment.
[0021] Further, the wellness analyzer 110 characterizes a patient
as it determines the wellness output 116. In this manner, the
wellness analyzer 110 creates an internal patient model or "virtual
patient." The virtual patient can then be tested, in a simulation
mode, with statistical inputs in lieu of, or in addition to, sensor
data 112 so as to determine a predictive wellness output 118.
Advantageously, predictive wellness 118 indicates how a patient is
likely to respond to various physical, medical and environmental
conditions so as to reveal physiological strengths and weaknesses.
For example, predictive wellness 118 may reveal a risk of an
immediate deterioration in health due to, for example, septic
shock, infection, embolism, pneumonia and pharmaceutical
side-effects, to name a few. Predictive wellness 118 may also
reveal longer-term susceptibility to particular diseases or disease
states, such as thrombosis, certain cancers, diabetes and heart
disease, as a few examples.
[0022] FIG. 2 illustrates a wellness analyzer embodiment 200 having
a wellness monitor 210 and one or more adjunct layers 220-260
coupled to the wellness monitor. The wellness monitor 210 processes
real-time data 280, such as from sensors 112 (FIG. 1). The adjunct
layers 220-260 each process non-real-time data 290, such as from
hospital records 120 (FIG. 1) and other databases 140 (FIG. 1). In
an embodiment, the adjunct layers 220-260 are specialized according
to the data source, such as a lab work layer 220, a pharmaceutical
or medications layer 230, a patient history layer 240, a genetics
layer 250 and an environmental layer 260, to name a few. The
wellness analyzer layers 210-260 extract features from the
real-time data 280 and non-real-time data 290 so that the wellness
monitor 210 may derive a wellness output 216, as described
above.
[0023] As shown in FIG. 2, the wellness analyzer layers 210-260
also store an internal analyzer state 272, including extracted
features and traces of the feature extraction process. The analyzer
state 272 may include the "virtual patient" characterization
described above. A simulator 270 layer utilizes the analyzer state
272 and, perhaps, the input data 280, 290 to generate statistical
data 274 to the wellness monitor 210, so as to generate a
predictive wellness output 218, also described above. The
predictive wellness output 216 may also include the internal
analyzer state 272 so as to determine what led to the
prediction.
[0024] FIG. 3 illustrates a wellness analyzer 300 embodiment having
a monitor layer 310, a labwork layer 330 and a pharmaceutical layer
350 among other layers not shown. The monitor layer 310 has sensor
301 inputs, logic blocks 320 and a wellness indicator 309 output.
In this particular embodiment, the logic blocks 320 are organized
into three levels including parameter 322, system 324 and
diagnostic 328 levels. The parameter "P" logic blocks 322 process
sensor data so as to derive parameters 312, which are numerical
variables and, perhaps, waveforms indicative of the status of
various physiological subsystems. For example, an optical sensor
attached to a fleshy tissue site can generate numerical parameters
such as oxygen saturation, pulse rate and total hemoglobin among
others, as well as a plethysmograph waveform, which are related to
arterial and/or venous blood flow at a particular body location.
This information, in turn, provides some information regarding the
body's circulatory system. Although selected parameters can be
displayed by the monitor, at this level most of the derived
information is internal to the machine.
[0025] The system "S" logic blocks 324 are responsive to parameter
levels, slopes, trends, variability, patterns and waveform
morphology, instantaneously and over time, so as to provide system
indicators 314, which describe some aspect of the physiological
status of one or more of the biological systems enumerated above,
including the circulatory, respiratory and neurological systems, to
name a few. For example, physiological system indicators may vary
from a specified number of measured events per unit time to a
general measure of physiological system health. These physiological
system indicators are vaguely analogous to blood chemistry results
from a lab, i.e. an indicator may be a measured number, such as
"significant desaturations per minute" along with a specified
acceptable or normal range. As such, a system indicator is
typically available for a nurse or doctor's review, but does not
provide an overall diagnosis of a patient condition. The system
indicators 314 are, in turn, input to one or more diagnostic logic
blocks 328.
[0026] The diagnostic "D" logic blocks 328 generate a wellness
output 309, i.e. one or more decisions regarding a person's health
that are intended as an ultimate diagnosis for a caregiver's
evaluation. A wellness output 309 can range from a relatively mild
diagnosis, such as "occasional atrial arrhythmias" to significant
health concerns, such as "apnea" to immediate and life threatening
concerns, such as "congestive heart failure." Further examples are
described below with respect to FIG. 4. In other embodiments,
different types, numbers and organizations of logic blocks may be
utilized in the monitor layer 310.
[0027] Also shown in FIG. 3, other wellness analyzer layers 330,
350 also have logic blocks that feed into the monitor layer 310
and/or other layers. For example, the labwork layer 330 may have
system level logic blocks that determine where a measured value,
such as a particular blood factor, falls relative to a normal range
of values. That result would then input into a monitor layer 310
diagnostic block 328, which combined with sensor derived
measurements would yield a particular diagnosis. As another
example, the pharmaceutical layer 350 may have diagnostic level
logic blocks that compare known side-effects of currently
prescribed drugs with one or more system-level logic block outputs
314 on the monitor layer 310 so as to generate an input to a
diagnostic-level logic block 328 on the monitor layer 310. In this
manner, the monitor 310 takes into account known drug side-effects
in determining the wellness output 309. In various embodiments,
there may be a back and forth flow of information 360 between
adjunct layers 330, 350 and between the monitor layer 310 and the
adjunct layers 330, 350 involving one or more levels of logic
blocks in each layer.
[0028] FIG. 4 shows an illustrative wellness monitor embodiment 400
having sensor 401 inputs, logic blocks 403 and a wellness indicator
409 output. The logic blocks 403 include parameter 404, system 406
and diagnostic 408 blocks. For example, at the parameter level 404,
an optical sensor 412 generates an absorption plethysmograph from
which a first parameter block 420 generates blood oxygen saturation
422 and pulse rate 424. An acoustic sensor 414 generates breathing
sound waveforms from which a second parameter block 430 generates a
respiration rate 432. A temperature sensor generates a body
temperature measurement 416. A blood pressure cuff 418 generates a
pressure plethysmograph from which a third parameter block 440
generates systolic and diastolic blood pressure measurements
442.
[0029] At the (physiological) system level 406, a first system
block 450 inputs oxygen saturation 422 and pulse rate 424 and
generates a circulatory system output 452. For example, the first
system block 450 may indicate at a particular point in time that
oxygen saturation 422 is falling in conjunction with a relatively
steady pulse rate. A diagnostic logic block 480 responds to the
circulatory system output 452 to generate a wellness state output
409 comprising a "cardiac distress" message. Details of the message
alert the caregiver that something is wrong with the patient's
heart as it is unable to compensate a drop in oxygen saturation
with an increase in circulatory system blood flow.
[0030] Also at the system level 406, a second system block 460
inputs oxygen saturation 422 and respiration rate 432 and generates
a respiratory system output 462. For example, the second system
block 460 may indicate at a particular point in time that oxygen
saturation 422 is falling in conjunction with a rising respiration
rate. A diagnostic logic block 480 responds to the respiratory
system output 462 to generate a wellness state output 409
comprising a "respiratory pathway blockage" message. Details of the
message alert the caregiver that a rising respiration rate is not
resulting in increased oxygen delivery to the lungs.
[0031] Further at the system level 406, a third system block 470
inputs body temperature 416 and blood pressure 442 and generates a
circulatory system output 472 related to hemodynamic stability. The
third system block 470 may also input pulse rate 424 and
respiration rate 432 to generate a multiple system output 472
responsive to these four significant vital signs.
[0032] As described with respect to FIGS. 1-4, above, the wellness
analyzer 110 (FIG. 1) can range from a general purpose computer to
a special-purpose signal processor or from a processor array to a
distributed network of computers or other processing devices. Also
as described above, the wellness monitor 210 (FIG. 2) and adjunct
layers 220-260 (FIG. 2) can be implemented in hardware, software,
firmware or a combination of these. In an embodiment, the wellness
analyzer 110 (FIG. 1) is a single instrument having plug-ins, one
or more displays, keyboards, various standard communications
interfaces, one or more signal processors and an instrument
management processor. In an embodiment, the various processing
layers 210-260 (FIG. 2) are implemented in software and/or
firmware. In an embodiment, the various logic blocks 320 (FIG. 3)
are subprograms or subroutines or otherwise identifiable portions
within software and/or firmware.
[0033] FIG. 5 illustrates details of a parameter logic block 500.
As described with respect to FIG. 3 (322) above, a parameter logic
block has a sensor input 501 and parameter 509 output. The sensor
input 501 can be any of a variety of sensor signals such as
photodiode detector current generated by an optical sensor,
piezoelectric current from an acoustic sensor or electrode voltage
from an ECG sensor, to name just a few. A hardware interface 510
apart from the parameter block 500 drives the sensor, if necessary,
and conditions, samples and digitizes the sensor input 501 into
sensor data 512. A signal extraction process 520 extracts one or
more physiological signals 525 from the sensor data 512. For
example, signal extraction for a pulse oximetry sensor includes
demodulating the red and IR signal components. The sensor signal(s)
525 are then analyzed 530 so as to derive physiological parameters
509. For pulse oximetry, signal analysis 530 would include deriving
a red over IR ratio and using that ratio with a lookup table to
calculate an oxygen saturation parameter.
[0034] FIG. 6 illustrates details of a system logic block 600. As
described with respect to FIG. 3 (324) above, a system logic block
600 has parameter inputs 601 as well as, perhaps, sensor data
inputs. The system logic block 600 also has a system status 609
output. In an embodiment, a system logic block 600 has an input
selector 610, a feature extractor 620, feature storage 630 and a
feature analyzer 640. A controller 390 (FIG. 3) provides control
inputs (not shown) to each of the input selector 610, feature
extractor 620, feature storage 630 and feature analyzer 640, as
described in further detail below. In an embodiment, the feature
extractor 620 has level 622, trend 624, pattern 625, statistics 627
and morphology 629 functions, as a few examples.
[0035] As shown in FIG. 6, the input selector 610 determines which
parameter blocks 322 (FIG. 3) feed its particular system block 324
(FIG. 3). Further, the input selector 610 routes the selected
parameters to specific feature extractor functions. The level
function 622 is responsive to input parameters 603 rising above or
falling below a predetermined threshold. The trend function 624 is
responsive to input parameters 603 having a positive or negative
rate of change above a predetermined absolute value over a
predetermined time interval. The pattern function 625 is responsive
to a predetermined parameter behavior over a predetermined time
interval, such as threshold crossings per minute. The statistics
function 627 determines parameter behavior over a predetermined
sample size, such as mean, variance and correlation (with other
parameters) to name a few. The morphology function 629 analyzes
sensor waveform characteristics, such as plethysmograph dicrotic
notch behavior, as one example. Other feature extraction functions
not shown may include an FFT function, to determine frequency
characteristics such as detection of sensor waveform modulation,
among others. Using FIG. 4, described above, as a system logic
block example, the input selector of the first system block 450
(FIG. 4) routes both the SpO.sub.2 parameter and the PR parameter
to the trend 624 function, so that the system block output 452
(FIG. 4) is responsive to upward or downward trends of both
SpO.sub.2 and pulse rate.
[0036] Further shown in FIG. 6, the feature analyzer 640 inputs the
extracted features 607 so as to indicate the functioning of a
physiological system, such as the circulatory, respiratory,
neurological and other systems cited above. In an embodiment, the
status output 609 is a list of status codes indicating a particular
physiological system is okay or is in various degrees of distress.
In an embodiment, the feature analyzer 640 accesses a look-up table
650 to generate the status codes according to extracted features.
In an embodiment, the status codes are prioritized according to
severity. In an embodiment, the look-up table 650 is uploaded from
the controller 390 (FIG. 3) according to particular system logic
block inputs and functions.
[0037] Also shown in FIG. 6, feature storage 630 stores traces of
selected input parameters along with extracted parameter features.
In this manner, each of the system blocks 600 builds a
characterization of the patient with respect to the particular
physiological system monitored. The sum of this patient
characterization across all system blocks creates a "virtual
patient" that can be tested by a simulator 870 (FIG. 8).
[0038] During simulation, independent parameters 601 and,
potentially, some sensor data are simulated and dependent
parameters are "played-back" from feature storage 630 accordingly.
The input selector 610 combines the simulated parameters 601 and
the playback parameters 632 as inputs to the feature extractor 620.
The feature analyzer 640 then responds to the simulated extracted
features 607 so as to determine a corresponding system status 609
according to how the patient historically responds. Simulation is
described in further detail with respect to FIG. 8, below.
[0039] FIG. 7 illustrates details of a diagnostic logic block 700
having system status 701 inputs and a wellness output 703. The
diagnostic block 700 comprises an expert system 710, a diagnostic
knowledgebase 720 and an output generator 730. The expert system
710 receives the system status 701 inputs, which are generated by
the various system logic blocks 600 (FIG. 6). In an embodiment,
system status 701 is input as status codes, as described above. The
expert system 710 compiles the status codes and interprets those
codes according to the diagnostic knowledge base 720 so as to
generate a diagnostic output 712. In an embodiment, the diagnostic
output 712 is one or more diagnostic codes, which are a compiled
diagnostic interpretation of the system status codes. The output
generator 730 determines the form and format of the wellness output
703 according to the diagnostic codes 712. In an embodiment, the
wellness output 703 is any or all of a wellness index 732,
diagnostic message(s) 734 and alarms and displays 738. In a
predictive wellness mode, the predictive knowledge base 740 is used
in lieu of the diagnostic knowledge base 720, as described with
respect to FIG. 8, below.
[0040] In an embodiment, a wellness index 732 is a numerical,
alphanumeric, color or other scale or combination of scales that
embodies the sum total of a diagnostic output 712. For example, a
wellness index of 10 indicates no significant problems or issues
reported by any system block; a wellness index of 8 indicates a
minor problem reported by at least one system block; a wellness
index of 6 indicates a significant problem reported by at least one
system block or minor problems reported by multiple system blocks;
a wellness index of 4 indicates significant problems reported by
multiple system blocks; and a wellness index of 2 indicates a major
problem reported by at least one system block.
[0041] In an embodiment, a diagnostic message 734 is natural
language text indicating a clinical decision in response to system
status codes 701. For instance, a diagnostic message might state a
potential cause for specific symptoms such as "patient displaying
symptoms of apnea." In an embodiment, the diagnostic message 734
might also suggest a remedy for the symptoms or a possible course
of treatment. indicators that alone or combined with other wellness
state outputs 732, 734 function to communicate specifics regarding
patient health or illness to a caregiver. Alarms include single
frequency, mixed frequency and varying frequency sounds. Displays
include text; bar, Cartesian, polar or 3-D coordinate graphs; and
icons to name a few.
[0042] FIG. 8 illustrates a predictive wellness analyzer 800 having
a monitor 810, one or more adjunct layers 830, 850 and a simulator
870. The monitor 810 has parameter logic blocks 804, system logic
blocks 806, one or more diagnostic logic blocks 808, a controller
807 and a predictive wellness output 809. In an embodiment, the
predictive wellness analyzer 800 utilizes the same
hardware/software resources as the wellness analyzer 300 (FIG. 3).
That is, in a predictive wellness mode, the controller 807 disables
one or more of the sensor inputs 801 and/or parameter logic blocks
804 and enables the simulator 870. The controller 807 also
configures the system logic blocks 806 to access stored features
630 (FIG. 6) and configures the diagnostic block 700 (FIG. 7) to
utilize the predictive knowledgebase 740 (FIG. 7) in lieu of the
diagnostic knowledgebase 720 (FIG. 7). So configured, the
predictive wellness monitor 810 advantageously tests a virtual
patient constructed from a history of patient monitoring and
information provided by adjunct layers 830, 850 to assess patient
near-term health risks and, perhaps, longer term susceptibility to
disease and other health threats, as described above.
[0043] As shown in FIG. 8, the simulator 870 synthesizes one or
more sensor inputs 801 and/or one or more parameter 804 outputs.
For example, the simulator 870 may generate a first parameter block
812 output according to a first probability distribution 813;
disable a second parameter block 814; and generate an nth parameter
block 818 output according to an nth probability distribution 819.
In an example based upon FIG. 4, oxygen saturation 422 (FIG. 4) is
varied, say, between 85-95% and pulse rate 424 (FIG. 4) is varied,
say, between 100-150 bpm. Accordingly, respiration rate 432 (FIG.
4), temperature 416 (FIG. 4) and blood pressure 442 (FIG. 4) are
allowed to vary as dependent parameters based upon the simulated
blood pressure and respiration rate. That is, the non-simulated
parameters vary according to the historical (stored) patient
responses to like variations in oxygen saturation and pulse rate.
In this manner, the likely response of a patient's respiratory
system and hemodynamics to normal variations in oxygen delivery are
measured.
[0044] Further shown in FIG. 8, the simulated parameters 813, 819
feed the system logic blocks 806. In the predictive mode, the
controller 807 alters the system block 806 pathways so that the
system status outputs 805 are responsive only to the simulated
parameters 813, 819 and the stored features 632 (FIG. 6) recorded
by the feature extractor 620 (FIG. 6) in the wellness mode. Hence,
the extracted features of simulated (independent) parameters
trigger the recall of dependent parameter features based upon
monitored history. Accordingly, the system blocks 806 generate a
system status 805 that reflects both the simulated parameter inputs
813, 819 and the historical patient response to those
parameters.
[0045] Continuing the example based upon FIG. 4, as oxygen
saturation and pulse rate are generated by the simulator 870, the
feature storage of the system blocks 806, in "playback" mode,
internally generates the corresponding "predicted" respiration
rate, temperature and blood pressure parameters. The corresponding
system status 805 is input to the diagnostic block 808, which
generates a predictive wellness 809 output, which differs from the
wellness output 703 (FIG. 7). In an embodiment, instead of a
clinical diagnosis of an existing condition based upon the
diagnostic knowledge base 720 (FIG. 7), predictive wellness
indicates the likelihood of a near-term deterioration of an
existing condition, such as a risk of infection, septic shock or an
adverse drug reaction, to name a few, according to the predictive
knowledge base 740 (FIG. 7).
[0046] A wellness analysis system has been disclosed in detail in
connection with various embodiments. These embodiments are
disclosed by way of examples only and are not to limit the scope of
the claims that follow. One of ordinary skill in art will
appreciate many variations and modifications.
* * * * *