Patent Application
20240057900
The present invention relates generally to wearable devices, and more particularly to wearable biometric sensor technology for physiological monitoring for medical, health, and fitness applications.
The “holy grail” of diabetes management would comprise a truly noninvasive, continuous blood glucose monitoring solution that would be completely painless and nearly invisible to the end user. Many attempts have been made to provide such a commercial solution, but none have succeeded.
Continuous blood glucose monitoring solutions are commercially available in the marketplace today, such as the Dexcom G6 CGM system (Dexcom, Inc., San Diego, California). These conventional monitoring systems take regular samples of body fluids (such as interstitial fluid), via microneedles or other minimally invasive modalities, and estimate blood glucose from sensor readings. However, by nature they are at best minimally invasive, and typically cause agitation to the underlying skin. Moreover, the concentration of glucose in interstitial fluid typically lags that of blood by several minutes, which may delay urgent feedback to the end user. Of equal importance, the form-factor of conventional glucose monitoring systems is typically that of a patch, which may feel awkward to many end users.
As reported by John Smith in “The Pursuit of Noninvasive Glucose: Hunting the Deceitful Turkey”, a myriad of researchers have labored over a noninvasive method of noninvasively measuring blood glucose levels accurately enough to dose insulin or glucose. However, the results have not been suitable for commercial use. A novel approach to this problem is needed.
It should be appreciated that this Summary is provided to introduce a selection of concepts in a simplified form, the concepts being further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of this disclosure, nor is it intended to limit the scope of the invention.
According to some embodiments of the present invention, a method of generating a blood glucose estimation for a subject includes the following steps performed by at least one processor: receiving real-time PPG data from a PPG sensor attached to the subject, and generating a blood glucose estimation for the subject via an adaptive predictive model using the real time PPG data. Exemplary adaptive predictive models include, but are not limited to, regression models, machine learning models, and classifier models. Generating the blood glucose estimation may include generating a current blood glucose estimation. Generating the blood glucose estimation may include generating a future blood glucose estimation. In some embodiments, a current blood glucose estimation and a past blood glucose estimation may be processed to predict a future blood glucose estimation.
The method may further include receiving a measurement of blood glucose from a blood glucose monitoring device and, in response to receiving the measurement of blood glucose, updating one or more parameters of the adaptive predictive model in real-time to improve blood glucose estimation accuracy of the adaptive predictive model. In some embodiments, the measurement of blood glucose is a real-time measurement.
The method may further include receiving subject activity information from a motion sensor attached to the subject, receiving a real-time measurement of blood glucose from a blood glucose monitoring device in response to the subject activity information and, in response to receiving the real-time measurement of blood glucose, updating one or more parameters of the adaptive predictive model in real-time to improve blood glucose estimation accuracy of the adaptive predictive model.
The method may further include detecting whether the generated blood glucose estimation is above or below a threshold. In response to determining that the generated blood glucose estimation is above or below the threshold, the method further includes receiving a measurement of blood glucose via a blood glucose monitoring device, and updating one or more parameters of the adaptive predictive model in real-time to improve blood glucose estimation accuracy of the adaptive predictive model.
In some embodiments, the method further includes sending an alert to a remote device that the generated blood glucose estimation is above or below a threshold.
In some embodiments, the at least one processor is located within a wearable device worn by the subject. The wearable device may be configured to be worn at an ear of the subject, on a limb of the subject, as a patch attached to the subject, or on a digit of the subject.
In some embodiments, the wearable device includes the PPG sensor.
In some embodiments, the at least one processor is located within a wearable device worn by the subject, and the wearable device includes the PPG sensor and the blood glucose monitoring device.
In some embodiments, the PPG sensor is an imaging sensor.
According to other embodiments of the present invention, a wearable device includes a PPG sensor and at least one processor configured to generate a blood glucose estimation for a subject wearing the wearable device via an adaptive predictive model using real-time PPG data from the PPG sensor. The wearable device may be configured to be worn at an ear of the subject, on a limb of the subject, as a patch attached to the subject, or on a digit of the subject. Exemplary adaptive predictive models may include, but are not limited to, regression models, machine learning models, and classifier models.
The at least one processor may be configured to receive a measurement of blood glucose from a blood glucose monitoring device and, in response to receiving the real-time measurement of blood glucose, update one or more parameters of the adaptive predictive model in real-time to improve blood glucose estimation accuracy of the adaptive predictive model. The at least one processor may be configured to receive a measurement of blood glucose from a blood glucose monitoring device and, in response to determining that the generated blood glucose estimation is above or below a threshold, update one or more parameters of the adaptive predictive model in real-time to improve blood glucose estimation accuracy of the adaptive predictive model. The at least one processor may be configured to send an alert to a remote device that the generated blood glucose estimation is above or below a threshold.
In some embodiments, the PPG sensor is an imaging sensor.
According to other embodiments of the present invention, a method of improving blood glucose estimation accuracy of an adaptive predictive model (e.g., a regression model, a machine learning model, a classifier model, etc.) includes the following steps performed by at least one processor: a) receiving, within a receiving period, real-time PPG data from a PPG sensor attached to a subject and a blood glucose measurement from a blood glucose monitoring; b) generating features from the received PPG data; c) storing the features and the blood glucose measurement; and d) updating one or more parameters of the adaptive predictive model in real-time by processing the stored features in context with the stored blood glucose measurement, wherein the updated one or more parameters improves the blood glucose estimation accuracy of the adaptive predictive model. The features and the blood glucose measurement may be stored in a data buffer, such as a FIFO (first-in-first-out) buffer, although other types of data buffers may be utilized. Steps a)-d) may be repeated over one or more subsequent time periods to improve estimation accuracy of the model.
The method may include generating a blood glucose estimation for the subject via the adaptive predictive model, then determining whether the generated blood glucose estimation is above or below a threshold. Responsive to determining that the generated blood glucose estimation is above or below a threshold, another measurement of blood glucose via the blood glucose monitoring device is received, and the one or more parameters of the adaptive predictive model are updated in real-time.
In some embodiments, generating features from the received PPG data includes generating features at feature generation intervals within the receiving period via a sliding time window. In some embodiments, updating the one or more parameters of the adaptive predictive model further includes processing the stored blood glucose measurement and a previously stored blood glucose measurement, and generating an interpolation between the stored blood glucose measurement and the previously stored blood glucose measurement. Processing the stored blood glucose measurement and the previously stored blood glucose measurement may include processing a plurality of previously stored blood glucose measurements. Processing the stored blood glucose measurement and the previously stored blood glucose measurement may include generating an interpolation of expected blood glucose measurements.
In some embodiments, the PPG sensor is an imaging sensor.
In some embodiments, processing the stored features in context with the stored blood glucose measurement includes processing a function of at least one of the stored features.
In some embodiments, processing the stored features in context with the stored blood glucose measurement includes calculating statistical information for a temporal sequence of at least one of the stored features. In some embodiments, processing the stored features in context with the stored blood glucose measurement includes calculating statistical information for a plurality of temporal sequences of at least one of the stored features.
In some embodiments, processing the stored features in context with the stored blood glucose measurement includes calculating weighted statistical information for a plurality of temporal sequences of at least one of the stored features.
According to other embodiments of the present invention, a system for improving blood glucose estimation accuracy of an adaptive predictive model (e.g., a regression model, a machine learning model, a classifier model, etc.) includes at least one processor configured to: receive, within a receiving period, real-time PPG data from a PPG sensor attached to a subject and a blood glucose measurement from a blood glucose monitoring device; generate features from the received PPG data; store the features and the blood glucose measurement; and update one or more parameters of the adaptive predictive model in real-time by processing the stored features in context with the stored blood glucose measurement, wherein the updated at least one parameter improves blood glucose estimation accuracy of the adaptive predictive model.
In some embodiments, the at least one processor is further configured to: generate a blood glucose estimation via the adaptive predictive model; determine whether the generated blood glucose estimation is above or below a threshold; and in response to determining that the generated blood glucose estimation is above or below the threshold, update the one or more parameters of the adaptive predictive model in real-time. The at least one processor may be configured to send an alert to a remote device that the generated blood glucose estimation is above or below the threshold.
It is noted that aspects of the invention described with respect to one embodiment may be incorporated in a different embodiment although not specifically described relative thereto. That is, all embodiments and/or features of any embodiment can be combined in any way and/or combination. Applicant reserves the right to change any originally filed claim or file any new claim accordingly, including the right to be able to amend any originally filed claim to depend from and/or incorporate any feature of any other claim although not originally claimed in that manner. These and other objects and/or aspects of the present invention are explained in detail below.
The accompanying drawings, which form a part of the specification, illustrate various embodiments of the present invention. The drawings and description together serve to fully explain embodiments of the present invention.
The term “subject”, as used herein, typically refers to a human being in context of the invention description. However, in context of the invention, a subject may also be a living creature that is not a human being.
The term “biometric” generally refers to a metric for a subject generated by processing physiological (i.e., biological) information from the subject. Nonlimiting examples of biometrics may include: heart rate (HR), heart rate variability (HRV), RR-interval, respiration rate, weight, height, sex, physiological status, overall health status, disease conditions, injury status, blood pressure, arterial stiffness, cardiovascular fitness, VO2max, gas exchange analysis metrics, blood analyte levels fluid metabolite levels, and the like.
The terms “biometric” and “physiological metric”, as used herein are interchangeable.
The term “real-time” is used herein to describe a process that requires a period of time that appears substantially real-time to a human individual. Thus, the term “real-time” is used interchangeably to mean “near real-time” or “quasi-real-time”. Namely, a “real-time” process may refer to an “instantaneous process” but may also refer to a process that generates an output within a short enough processing time to (in effect) be as useful as an instantaneous process (in context of a particular use case). For example, in practicality, a process that requires several seconds or minutes to generate a blood pressure metric for a subject may be considered to be a real time process, as used herein, even though blood pressure may be changing each second, as the use case may involve a sedentary state for the subject where subtle changes in blood pressure may be insignificant and averaged out.
The terms “respiration rate” and “breathing rate”, as used herein, are interchangeable.
The terms “heart rate” and “pulse rate”, as used herein, are interchangeable.
The term “system”, as used herein, refers to a collection of physical and/or computational materials that may be unified by a common function.
The terms “motion sensor”, as used herein, refers to a sensor configured to sense motion information (e.g., from a subject). Nonlimiting examples of motion sensors may comprise: single- or multi-axis inertial sensors (such as accelerometers, gyroscopes, MEMS motion sensors, and the like), optical scatter sensors, blocked channel sensors, and the like.
The term “photoplethysmography” (PPG), as used herein, refers to a method of generating physiological information from PPG waveforms collected via a PPG sensor.
The term “PPG waveform”, as used herein, refers to physiological waveform data resulting from a temporal modulation of photon flux through physiological material.
The term “PPG sensor”, as used herein, refers to a sensor configured to sense photons and generate PPG waveform data. A typical PPG sensor may comprise an optical sensor configured to sense photons in the optical spectrum (i.e., an electromagnetic wavelength range of ˜10 nm to 103 μm, or electromagnetic frequencies in the range from ˜300 GHz to 3000 THz). Nonlimiting examples of optical sensors may comprise inorganic and/or organic photodetectors (such as photoconductors, photodiodes, phototransistors, phototransducers, and the like), reverse-biased light-emitting diodes (LEDs) or other reverse-biased optical emitters, imaging sensors, photodetector arrays, and the like. Additionally, a typical PPG sensor may also comprise a photon (photonic) emitter to generate a photon flux through a physiological pathway. However, in some cases, ambient photons or photons from another source (that is not part of the PPG sensor) may be used to generate photons. Typical PPG sensors may comprise photon emitters that are optical emitters, such as inorganic and/or organic light-emitting diodes (LEDs), laser diodes (LDs), microplasma sources, or the like. PPG sensors may also comprise a motion sensor for the purposes of generating subject activity data and/or providing a noise reference for attenuating motion artifacts in PPG waveform data.
The terms “sensor”, “sensing element”, and “sensor module”, as used herein, are interchangeable and refer to a sensor element or group of sensor elements that may be utilized to sense information, such as information (e.g., physiological information, body motion, etc.) from the body of a subject and/or environmental information in a vicinity of the subject. A sensor/sensing element/sensor module may comprise one or more of the following: a detector element, an emitter element, a processing element, optics, or optomechanics, sensor mechanics, mechanical support, supporting circuitry, and the like. Both a single sensor element and a collection of sensor elements may be considered a sensor, a sensing element, or a sensor module. A sensor/sensing element/sensor module may be configured to both sense information and process that information into one or more metrics.
As used herein, the term “processor” broadly refers to a signal processing circuit or computing system, or a computational method, which may be localized and/or distributed. For example, a localized signal processing circuit may comprise one or more signal processing circuits or processing methods localized to a general location, such as to a wearable biometric monitoring device. Examples of such devices may comprise, but are not limited to, an earpiece, a headpiece, a finger clip, a toe clip, a limb band (such as an arm band or leg band), an ankle band, a wrist band, a digit (e.g., finger or toe) band, a nose band, a sensor patch, jewelry, a patch, apparel (clothing) or the like. Examples of a distributed processing circuit include “the cloud,” the internet, a remote database, a remote processor computer, a plurality of remote processing circuits or computers in communication with each other, etc., or processing methods distributed among one or more of these elements. The difference between distributed and localized processing circuits is that a distributed processing circuit may include delocalized elements, whereas a localized processing circuit may work independently of a distributed processing system. Microprocessors, microcontrollers, or digital signal processing circuits represent a few non-limiting examples of signal processing circuits that may be found in a localized and/or distributed system.
The terms “mobile application”, “mobile app” and “app”, as used herein, are interchangeable and refer to a software program that can run on a computing apparatus, such as a mobile phone, digital computer, smartphone, database, cloud server, processor, wearable device, or the like.
The term “health”, as used herein, is broadly construed to relate to the physiological status of an organism or of a physiological element or process of an organism. For example, cardiovascular health may refer to the overall condition of the cardiovascular system, and a cardiovascular health assessment may refer to an estimate of blood pressure, VO2max, cardiac efficiency, heart rate recovery, arterial blockage, arrhythmia, atrial fibrillation, or the like. A “fitness” assessment is a subset of a health assessment, where the fitness assessment refers to how one's health affects one's performance at an activity. For example, a VO2max test can be used to provide a health assessment of one's mortality or a fitness assessment of one's ability to utilize oxygen during an exercise.
The term “blood pressure”, as used herein, refers to a measurement or estimate of the pressure associated with blood flow of a person, such as a diastolic blood pressure, a systolic blood pressure, a mean arterial pressure, pulse pressure, or the like. The blood pressure may be with reference to any location on the body where blood vessels and blood flow exists (i.e., brachial, thoracic, subclavian, femoral, tibial, radial, carotid, and the like). The term “blood pressure” is abbreviated as “BP” throughout this document.
As used herein, any device or system is considered to be remote to another device or system as long as there is no physical connection between them. As a point of clarity, the term “remote” does not necessarily mean that a remote device is a wireless device or that it is a long distance away from a device in communication therewith. For example, in some cases, two devices may be considered remote devices with respect to each other even if there is a physical connection between them. In this case, the term “remote” is intended to reference a device or system that is distinct from another device or system or that is not substantially reliant on another device or system for core functionality. For example, a computer wired to a wearable device may be considered a remote device, as the two devices are distinct and/or not substantially reliant on each other for core functionality.
The terms “sampling frequency”, “signal analysis frequency”, and “signal sampling rate”, as used herein, are interchangeable and refer to the number of samples per second (or per other time unit) taken from a continuous sensor or sensing element (for example, the sampling rate of the thermopile output in a tympanic temperature sensor or the sampling rate of the PPG signal from a PPG sensor).
It should be noted that “algorithm” and “circuit” are referred to herein. An algorithm refers to a computational instruction set, such as an instruction set with sequential steps and logic, that may be in memory whereas a circuit refers to physical components and/or traces (or path) that may implement such logic operations in the digital, analog, and/or quantum domains. These circuits may typically comprise electric circuits but may alternatively comprise elements that are photonic, electromagnetic, magnetic, acoustic, quantum, or the like.
To address these limitations, methods and apparatus according to the present invention provide for continuously generating blood pressure estimates (and/or other biometric estimates including, but not limited to, EEG estimates, respiratory metric estimates, core body temperature estimates, blood glucose estimates, etc.) via a real-time adaptive predictive model. These methods and apparatus leverage continuous PPG measurements from a subject, combined with at least one BP (or other biometric) measurement from a subject, to update, in real-time, a predictive model for that subject that is more accurate in estimating BP (or other biometric) for that subject (than prior to the update). The methods of the present invention may be implemented in a computational system that is configured to receive the PPG and BP (or other biometric) data and process this data to improve estimation accuracy. Namely, the model may be configured to generate a BP estimate for a given set of PPG input features, such that the BP estimate is a function of the PPG features, and the parameters of the model may be updated over time as recurring BP measurements (e.g., from a cuff-based BP monitor) are processed to improve the error of the model. The PPG sensors may be wearable and thus integrated into an apparatus or material that is in proximity to the skin of the subject. Alternatively, the PPG sensors may be stand-off sensors, such as imaging sensors (e.g., cameras) remote scanning sensors (e.g., radar, doppler, etc.), or the like, as described later herein.
In some cases, the computational system may be worn as an ear-worn device (e.g., hearables/hearing aids) 10, as a limb-worn (e.g., wrist, arm, leg) device 12, as a patch 14, or as a finger clip 16, as illustrated in
A key benefit of the biometric estimation (over the biometric measurement) is that the estimation may be continuous and painless, whereas the biometric measurement may be discrete and irksome to measure (such as measuring blood pressure with an automated cuff-based BP monitor or measuring blood glucose with a blood sample from a finger prick). Thus, even though the biometric estimation may have less measurement acuity than the actual biometric measurement, the ability to provide a “good enough” estimate (in between actual measurements) may outweigh the downside of potentially lower acuity.
These wearable PPG devices 12-16 may be in communication (e.g., electrical, optical, or wireless) with a blood pressure monitoring device, such as a blood pressure cuff 18 (such as that shown on the arm of the subject wearing the PPG earpiece 12 in
In some embodiments of the present invention, referred to as an adaptation process, a plurality of BP measurements from a cuff-based BP monitor 18 or other BP monitoring device and PPG measurements are processed together to improve the accuracy of the BP estimation. Once the model has been autonomously optimized for the subject and updated (
Alternatively, BP measurements may be received and processed routinely, referred to as an augmentation process, such that the adaptive predictive model may be continuously augmented over time based on updated BP measurements (such as those taken from an automated cuff-based BP monitor). In augmentation, updating an adaptive predictive model according to embodiments of the present invention may be repeated continuously, several times an hour, with each new BP measurement update.
The triggering of a model update (
Triggering based on a predetermined timing may be set as fixed or as a user adjustable parameter. This paradigm may be particular useful for hospital use cases and similarly immobile use cases, where the subject is lying at rest (e.g., in a hospital bed).
Triggering based on motion may be achieved by sensing activity above or below a threshold via a motion sensor (e.g., an accelerometer, imaging system, or other motion sensing device or component). This type of trigger may be especially useful for ambulatory monitoring of a biometric. If the motion status shows that excessive activity has occurred or that the frequency of excessive activity has increased, the computational system 100 may be triggered for more frequent model updates and the biometric measurement device may be triggered for more frequent measurements. This may help assure that estimation accuracy remains preserved. In keeping, if the motion status shows that the subject is resting or that the frequency of excessive activity is sufficiently low, the computational system 100 may be triggered for less frequent model updates and the biometric measurement device may be triggered for less frequent measurements. Similarly, this triggering may be dependent on activity status as opposed to a motion threshold. For example, an autonomously determined activity status of “running” or “walking” may trigger more frequent model updates and biometric measurements, whereas an autonomously determined activity status of “resting” or “sitting” may trigger less frequent model updates and biometric measurements. Methods of determining activity status via accelerometer data or imaging data are well known to those skilled in the art, such as in U.S. Pat. No. 10,610,158, which is incorporated herein by reference in its entirety.
If a wearable device according to embodiments of the present invention is removed from the subject and then re-worn, or if the device is repositioned on the subject, the biometric measurement device and the computational system may be triggered to take another biometric measurement and update the biometric estimation model, respectively. This autonomous triggering may help reestablish biometric estimation accuracy in the case that the wearable device has been temporarily disturbed or decoupled from the body. The autonomous determination that a device has been removed, repositioned, or re-worn can be executed by processing PPG data to determine signal quality or to determine other biometric parameters that may change as a wearable PPG device is positioned differently along the body. Methods of autonomously determining how a wearable device is being worn via PPG and motion sensing have been previously described, for example, in U.S. Pat. Nos. 9,794,653, 10,003,882, 10,512,403, and 10,893,835, the contents of which are incorporated herein by reference in their entireties.
A change in physiological status may also be autonomously detected and used to trigger another biometric measurement and update the biometric estimation model. For example, PPG sensor data (or other biometric data) may be processed to generate an assessment of stress status, cardiac status, respiratory status, or the like, and this physiological status update may be used as an autonomous trigger to take another BP measurement and update the model parameters. As one specific example, heart rate variability data from a PPG sensor (or other suitable biometric sensor) may be processed to indicate that one's stress state has changed (e.g., that the stress has notably increased or decreased), and this may provide the autonomous trigger. As another specific example, data from a PPG sensor (or other suitable biometric sensor) may be processed to generate subject breathing information (such as breathing rate, breathing volume, or breathing regularity (periodicity)/irregularity (aperiodicity)), and this may provide the autonomous trigger. For example, a significant change in breathing rate, breathing volume, or breathing regularity may be provide the autonomous trigger. This may be especially important for the accuracy of the invention in a field environment, as the transfer function between PPG information and subject blood pressure may be dependent on breathing dynamics. Methods of autonomously determining a change in physiological status via wearable PPG have been previously described, for example, in: U.S. Pat. Nos. 8,157,730, 8,929,966, 9,427,191, 10,413,250, 10,893,835, and 11,058,304, the contents of which are incorporated herein by reference in their entireties.
Similarly, in the case that an error (such as an operational error code) is detected by the computational system, an automatic triggering of a biometric measurement and model update may commence. This may help assure that accurate monitoring is robust to operational glitches. As one specific example, if the computational system receives an error code that the BP monitoring autocuff device which feeds the PPG-BP model has stopped inflating, this may trigger a system reset followed by another BP measurement and another PPG-BP model update.
Referring to
Then to decrease BP, the subject was asked to relax by terminating the isometric leg press. BP measurements from the cuff-based BP monitor (presented as a thick vertical line L1, with the top point of the line L1 representing the subject systolic BP and the bottom point of the line L1 representing the subject diastolic BP) were received every 60-to-90 seconds and processed (by a computational system). During an initial calibration phase of approximately 300 seconds, multiple values from the cuff-based BP monitor were processed along with multiple PPG readings to generate multiple PPG estimates (presented as a thin vertical line L2, in the same formalism as the cuff-based BP monitor readings). However, these estimates were not reported to the user, as the parameters of the adaptive predictive model were updated during this calibration phase to increase model accuracy such that it would be equivalent to that of the cuff-based BP monitor by the end of the calibration phase.
Following the calibration phase, continuous BP estimates were generated without updating model parameters for each new BP measurement. Rather, the remaining cuff-based BP monitor measurements are shown along with PPG estimations simply to note the excellent tracking between the PPG model estimates and the cuff-based BP monitor measurements. It should be noted that although the PPG estimates shown in
The test sequence of
Referring to
It should be emphasized that this invention is not limited to PPG-based BP estimations but may also be applied towards other PPG-based biometric estimations. Moreover, other measurement modalities outside of BP measurements may be used as the benchmark and basis for updating the adaptive predictive model. Nonlimiting examples of such biometric measurements and respective biometric estimations may comprise measurements and estimates of: breathing (respiration) rate, heart rate, cognitive load, intent (e.g., to take a mental or physical action), cardiac output, cardiopulmonary functioning, a cardiac condition or disease state (such as an arrythmia, premature contraction of the heart, heart damage, heart disease, plaque build-up, and the like), gas-exchange dynamics, blood analyte constituents (e.g., blood glucose level, blood urea level, bilirubin level, cholesterol level, etc.), and the like. Additional examples of other measurement/estimation modalities include monitoring ECG, EEG, EMG, EOG, blood flow volume, chest impedance, auscultatory monitoring, arterial line data, bloodwork data, and the like. As a specific example, it may be desirable to monitor the EEG readings for a subject to study brain wave patterns during an activity or during sleep. However, EEG is notoriously uncomfortable to wear, especially for sleep, where it would be more desirable to monitoring one's EEG with a more comfortable technology, such as PPG. Thus, a set of EEG electrodes may be used to provide EEG measurement data to be fed into the adaptive predictive model, and a different sensor modality, such as PPG, may be monitored simultaneously and fed into the adaptive predictive model to create a model for estimating EEG via PPG data (without requiring EEG data). Namely, once the PPG model has adapted to the EEG data, estimating the real-time EEG for a subject may then commence using real-time PPG data alone (i.e., without EEG sensing required).
Referring to
Referring to
It is to be understood that the steps illustrated in
Referring to
If the biometric estimation is above or below a threshold (Block 224), a real-time measurement of the biometric is received by the computational system from a biometric monitoring device (e.g., a blood pressure cuff 18,
It should be noted that BP estimation does not have to fall outside of a range in order for a calibration cuff reading to be called for and then used to increase accuracy of the estimations. Estimated BP can be in a normal range and a subsequent cuff reading can still be used to refine the accuracy. The adaptive predictive model can be updated merely based on set timed cuff-based readings, without regard to BP values versus a threshold. Similarly, the adaptive predictive model may be updated due to sensed changes in activity level (e.g., sensing a change in body motion via an accelerometer) or due to other sensor readings.
A remote device may be a smartphone of a medical provider, a nurse's station in a medical facility, or any other device that can alert a medical person as to the condition of the subject. The alert may also be sent to the blood pressure monitoring device (e.g., the blood pressure cuff 18,
It should be noted that, although Block 224 is presented as a “threshold” decision, Block 224 may be replaced with conditional logic for determining that a model update should occur. For example, rather than threshold logic based on one biometric estimation, Block 224 may comprise logic to determine the existence of a thresholding pattern of a plurality of biometric estimations (e.g., that have been stored in memory). In one nonlimiting example, this pattern may comprise a series of consecutively high (above normal) or consecutively low (below normal) BP estimations, and the determination of this pattern may then trigger a model update. In another nonlimiting example, this pattern may comprise an average of multiple biometric estimations that is determined to be above or below normal; if the pattern is determined to exist, then a model update may be triggered. The methods illustrated in
Referring back to
The received PPG data is processed to generate a plurality of real-time PPG features (Block 202,
Generating features from the received PPG data may comprise generating features at feature generation intervals (time-points) t=ki within the receiving period via a sliding window of time Δtw (
It should also be noted that, prior to generating a BP estimate (or other biometric estimate), the PPG features (characteristic features) may be actively normalized (e.g., weighted), to help ensure smooth temporal tracking of PPG-based BP estimations (or other biometric estimations) with BP measurements (or other biometric measurements). One normalization approach is to process the statistics of the stored features (e.g., the prior stored PPG features in memory) and to normalize by these statistics. Normalization may be performed by processing historical data over a plurality of feature generation time-points, by generating statistics for the historical data and normalizing by these statistics. This normalization process may be updated with each new feature generation time point (e.g., t=ki of
The aforementioned feature statistics themselves may also be employed as features to an adaptive predictive model, according to embodiments of the present invention. This may help provide smoother tracking (e.g., of BP estimations vs. BP measurements).
It should be noted that, as part of (or prior to) feature generation, preprocessing of the received sensor information (e.g., the PPG sensor data) and/or the received biometric measurement data (e.g., the BP measurement data) may be required. Additionally, it may be important to qualify the received data to reject “bad” data, generate a confidence score for the data, identify “good” data, or to classify data for further processing. A variety of preprocessing methodologies for PPG data (including associated motion sensor data) have been previously published and are well known to those skilled in the art, including, but not limited to: U.S. Pat. Nos. 10,834,483, 10,798,471, 10,631,740, 10,542,893, 10,512,403, 10,448,840, 9,993,204, 10,413,250, and PCT Application No. US20/49229, all of which are incorporated herein by reference in their entireties. Both passive and active methodologies of removing subject motion noise may be employed. Moreover, it should be noted that the optimal preprocessing may be feature-dependent. For example, regarding PPG data, for spectral domain features it may be desirable to remove or attenuate the “DC component” (e.g., the non-pulsatile component) from the PPG signal before feature generation. However, the DC component may be important for other features (such as time-domain features), or the DC component may even be a feature in itself. It should also be noted that PPG sensor data may comprise subject motion data (as described earlier), and this motion data may be utilized to reduce motion artifacts from optical sensor readings. The motion sensor may be integral to, or collocated with, the PPG sensor. Motion sensor data may be processed as a feature as well.
Preprocessing of biometric measurement data may also be useful. For example, in a preferred use case, a BP measurement from a BP cuff may comprise a discrete value of systolic and diastolic BP measurements. In some use cases, this data may be available to the computational system through an API (application programming interface) or through an application-specific interface. However, in some use cases, the BP measurement data received by the computational system of
Referring to
Depending on the type of model used, the model parameters may be different. For example, in a regression model, the model parameters may comprise at least one coefficient to the regression model. Nonlimiting examples of suitable regression models may comprise: linear, SVM, random forest, neural network, decision trees, a combination of these models, and the like. Other types of models outside of regression models may also be utilized; as a nonlimiting example, a classifier may be utilized, or a combination of classification and regression (as may be utilized in a convolutional neural network (CNN)). Updating the model may comprise processing the characteristic features (e.g., normalized characteristic features) and a stored blood pressure measurement to generate updated model parameters that reduce the estimation error of the adaptive predictive model 300. For example, the regression model may be solved for the recent BP measurement (or biometric measurement) and then the model coefficients may be updated. Alternatively, or additionally, a gradient-based optimization approach may be employed (such as classical gradient descent, Adam, Momentum, AdaGrad, RMSProp, AMSgrad, or the like) to update model coefficients with each new BP measurement.
Updating the adaptive predictive model in real-time may comprise processing a recent stored blood pressure measurement (associated with timepoint t=uj) and a prior stored blood pressure measurement (associated with time-point t=uj-1). In one embodiment, this may comprise generating an interpolation of expected blood pressure measurements (i.e., a temporal interpolation) between blood pressure measurements collected over time, such as an interpolation between the recent stored blood pressure measurement and the prior stored blood pressure measurement (or a plurality of prior stored blood pressure measurements). A specific example can be summarized in context of
As summarized above, there are many model constructs that may be used to generate the biometric estimation, and the general formalism of the function used to generate the biometric estimation is presented in
For implementing the methods of
The computational system 100 may be implemented as a plurality of discrete components, a fully integrated system, or a mixture of both. As a specific example, the computational system 100 may comprise a fully integrated microprocessor, with computational instructions for executing the processing steps of
The memory M may hold various categories of software and data, such as computer readable program code PC and/or an operating system OS. The operating system OS may control operations of the processor P, a PPG sensor (e.g., 12-16,
Alternatively, the computational system 100 may comprise an analog circuit configured to process the steps through analog processes. As another example, the computational instructions may be executed as a software library executed via a computational system (such as a processor). As another example, the computational system may comprise neural circuitry or quantum computing. Traditional, quantum, or neural processors may be utilized, or a combination of each.
A variety of components for enabling the system 100 of
The system may comprise input/output lines (i.e., ports or buses) to communicate with other systems, for receiving and giving data from/to external systems. For example, the input/output lines may communicate with at least one external processor, computational system, or apparatus. In one specific embodiment, a biometric estimation generated may be digitized and made available to an external computational system via a digital bus 106. In another embodiment, the input/output lines may communicate with one or more transceivers for communicating wirelessly with an external system. A variety of electronic communication configurations are well known to those skilled in the art.
In the case where a BP estimation is generated by the computational system of
It should be noted that one form of external system data may comprise meta data for the subject, and this meta data may be useful in processing biometric estimations in accordance with embodiments of the present invention. Namely, the computational system 100 of
As noted earlier, the system of
Because PPG information comprises rich information regarding blood flow, blood pressure is just one of many real-time hemodynamic parameters that may be extracted via embodiments of the present invention. Namely, a variety of hemodynamic parameters may be estimated via embodiments of the present invention, such as (but not limited to): arterial blood pressure, mean arterial pressure, systolic pressure variation, pulse pressure variation, stroke volume variation, right arterial pressure, right ventricular pressure, pulmonary artery pressure, mean pulmonary artery pressure, pulmonary artery wedge pressure, left atrial pressure, cardiac output, cardiac index, stroke volume, stroke volume index, systemic vascular resistance, systemic vascular resistance index, pulmonary vascular resistance, pulmonary vascular resistance index, stroke work index, ejection fraction, and the like. The ability of embodiments of the present invention to estimate these parameters is contingent on the right measurement device. For example, accurately estimating real-time ejection fraction via embodiments of the present invention would require collecting time-correlated measurement data form an accurate benchmark device, such as an echocardiogram monitoring device.
As just one example, measurement data may comprise EEG measurements and the associated biometric estimation that is generated may be an estimation of an EEG assessment. Nonlimiting examples of EEG assessments may comprise alertness level, dominating pattern (e.g., alpha, beta, theta, or delta), subject intent, identification of an abnormality, identification of normality, brain disorder, sleepiness, wakefulness, or the like. In this case, at least some of the characteristic features generated from processing the PPG sensor would be given significantly more or less weight than those used in estimating blood pressure. This is because the physiological relationship between EEG and PPG is quite different than that of BP and PPG. For example, PPG features related to time-domain variations in PPG peak locations (such as heart rate variability, the temporal location of systolic and diastolic peaks, as well as the temporal location of the dicrotic notch) may be more tightly correlated with EEG features.
As another example, if the measurement data comprises a gas-exchange (respiratory) analysis measurement, the associated biometric estimation that is generated may comprise an estimation of a gas-exchange analysis measurement. Nonlimiting examples of gas-exchange analysis measurements may comprise: carbon dioxide, oxygen, arteriovenous oxygen difference, exercise oscillatory breathing, fraction of carbon dioxide or oxygen in expired air, expiratory volume, metabolic equivalent, maximum voluntary ventilation, oxygen-uptake efficiency slope, partial pressure of end-tidal carbon dioxide or end-tidal oxygen, carbon dioxide output, respiratory exchange ratio, minute ventilation, volume of dead space, ventilatory threshold, ventilatory equivalent of oxygen or carbon diode, oxygen update, or the like.
In a similar example, if the measurement data comprises arterial blood gas measurements, the associated biometric estimation that is generated may comprise an estimation of an arterial blood gas measurement. Nonlimiting examples of arterial blood gas measurements include: H+ or pH level, total CO2, O2 content, partial pressure of oxygen or carbon dioxide, HCO3− (bicarbonate), SBCe, base excess, arterial oxygenation, venous oxygenation, oxygen extraction, or the like. For the cases of generating a gas-exchange analysis estimation or an arterial blood gas estimation, it may be especially important to receive PPG data comprising simultaneous multiwavelength (MWL) data (such as streaming PPG data from a MWL PPG sensor), in order to generate a set of characteristic features for a plurality of wavelengths of light. This is because the PPG feature distribution may be dependent upon the wavelength of light used, and this distribution may modulate differently in time depending on the respiratory or blood gas parameter being monitored. As just one example, the relative amplitude of PPG signals for different wavelengths of light will modulate in a characteristic way for low vs. high levels of blood oxygenation.
In another example, if the measurement data of
In another example, if the measurement data of
As noted earlier, one type of PPG sensor that may be utilized in the system and methods of
There are important benefits that may be afforded by utilizing an imaging sensor as the PPG sensor. Simultaneous data may be obtained from multiple distinct body locations (i.e., to map-out biometric estimations throughout the body). Also, simultaneous data at multiple wavelengths of electromagnetic energy (whether the photons are in the visible range or otherwise) may be obtained. For example, embodiments of the present invention described herein, when utilized with an imaging sensor as the PPG sensor, may be used to continuously estimate blood pressure for a subject at multiple body locations simultaneously, with measurement data provided by at least one blood pressure monitor. In one embodiment, the subject may be wearing a blood pressure cuff at the arm (e.g., via a brachial blood pressure cuff), and this measurement data is then fed to the system and methods of
It should also be noted that the blood pressure measurement device itself may be a remote imaging sensor in this invention. And in such case the benefit of a wearable PPG sensor may be the ability to passively assess blood pressure (via the blood pressure estimation) following BP measurements taken via video (collected via the imaging sensor). Reversely, the blood pressure measurement device in this invention itself may be a wearable PPG sensor (i.e., that has been optimized for measuring or estimating BP) or a blood pressure cuff, and the biometric sensor may be an imaging sensor. In such case, a key benefit of this approach may be that a remote imaging system can continuously and passively (in the background without requiring a change in subject behavior) improve BP estimations for the subject as new data from the blood pressure measurements continues to improve the adaptive predictive model.
The invention may also be applicable for the case where the biometric sensor data is not PPG sensor data but rather a different sensor modality (or a combination of these modalities in a sensor fusion approach), such as sensor data that is electromagnetic, auscultatory, electrical, magnetic, mechanical, thermal, or the like. In such case, the more general invention is presented in
In some cases, the estimations generated by the adaptive predictive model may be estimations of what the model predicts the estimations truly are in real-time. However, in some cases, the estimations generated by the adaptive predictive model may be estimations of what the model predicts the estimations will be in the impending future. One method of achieving this is via an adaptive predictive model that is tuned to generate future biometric estimations (for a future point in time) given a set of real-time data as opposed to current biometric estimations. Another method of achieving this may be to generate current biometric estimations, using methods as those outlined above, but then applying an additional layer to the model such that trends in past and current biometric estimations are further processed to generate predictions of future biometric estimations. Other approaches may be utilized.
A key benefit of predicting future biometric estimations (as opposed to current biometric estimations) may be that subjects wearing embodiments of the present invention may be informed that future undesirable biometric values may be imminent, such that they may take prophylactic measures to prevent these undesirable values. For example, a subject managing diabetes may benefit from knowing that their blood glucose levels are about to rise/drop sharply, enabling them to take a prophylactic dose of insulin/glucose to prevent this negative outcome. Similarly, a subject managing hypertension may benefit from knowing that their blood pressure is about to rise/drop sharply, enabling them to medicate accordingly.
Example embodiments are described herein with reference to block diagrams and flow diagrams. It is understood that a block of the block diagrams and flow diagrams, and combinations of blocks in the block diagrams and flow diagrams, can be implemented by computer program instructions that are performed by one or more computer circuits, such as electrical circuits having analog and/or digital elements. These computer program instructions may be provided to a processor circuit of a general purpose computer circuit, special purpose computer circuit, and/or other programmable data processing circuit to produce a machine, such that the instructions, which execute via the processor of the computer and/or other programmable data processing apparatus, transform and control transistors, values stored in memory locations, and other hardware components within such circuitry to implement the functions/acts specified in the block diagrams and flow diagrams, and thereby create means (functionality) and/or structure for implementing the functions/acts specified in the block diagrams and flow diagrams.
These computer program instructions may also be stored in a tangible computer-readable medium that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable medium produce an article of manufacture including instructions which implement the functions/acts specified in the block diagrams and flow diagrams.
A tangible, non-transitory computer-readable medium may include an electronic, magnetic, optical, electromagnetic, or semiconductor data storage system, apparatus, or device. More specific examples of the computer-readable medium would include the following: a portable computer diskette, a random access memory (RAM) circuit, a read-only memory (ROM) circuit, an erasable programmable read-only memory (EPROM or Flash memory) circuit, a portable compact disc read-only memory (CD-ROM), and a portable digital video disc read-only memory (DVD/BlueRay).
The computer program instructions may also be loaded onto a computer and/or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer and/or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions/acts specified in the block diagrams and flow diagrams. Accordingly, embodiments of the present invention may be embodied in hardware and/or in software (including firmware, resident software, micro-code, etc.) that runs on a processor such as a digital signal processor, which may collectively be referred to as “logic”, “circuitry”, “a module”, “an engine” or variants thereof.
It should also be noted that the functionality of a given block of the block diagrams and flow diagrams may be separated into multiple blocks and/or the functionality of two or more blocks of the block diagrams and flow diagrams may be at least partially integrated. Finally, other blocks may be added/inserted between the blocks that are illustrated. Moreover, although some of the diagrams include arrows on communication paths to show a primary direction of communication, it is to be understood that communication may occur in the opposite direction to the depicted arrows.
The foregoing is illustrative of the present invention and is not to be construed as limiting thereof. Although a few exemplary embodiments of this invention have been described, those skilled in the art will readily appreciate that many modifications are possible in the exemplary embodiments without materially departing from the teachings and advantages of this invention. Accordingly, all such modifications are intended to be included within the scope of this invention as defined in the claims. The invention is defined by the following claims, with equivalents of the claims to be included therein.
This application claims the benefit of and priority to U.S. Provisional Patent Application No. 63/132,233 filed Dec. 30, 2020, the disclosure of which is incorporated herein by reference as if set forth in its entirety.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/US2021/065119 | 12/23/2021 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 63132233 | Dec 2020 | US |
