Method and apparatus for generating assessments using physical activity and biometric parameters
The methods and apparatuses presented herein determine and/or improve the quality of one or more physiological assessment parameters, e.g., response-recovery rate, based on biometric signal(s) and/or motion signal(s) respectively output by one or more biometric and/or motion sensors. The disclosed methods and apparatuses also estimate a user's stride length based on a motion signal and a determined type of user motion, e.g., walking or running. The speed of the user may then be estimated based on the estimated stride length.
1. An assessment generation system configured to determine a blood pressure of a user using a photoplethysmography (PPG) sensor and a motion sensor comprised in a device disposed proximate at least some skin of the user such that at least the PPG sensor contacts a user's skin, the assessment generation system comprising a processing circuit configured to:
screen motion data from the motion sensor to determine an integrity of PPG data from the PPG sensor;
buffer the PPG data from the PPG sensor, wherein at least some of the buffered PPG data has integrity as determined responsive to the motion data;
generate a PPG waveform from the buffered PPG data; and
process the PPG waveform using a machine learning model to generate the blood pressure of the user.
2. The assessment generation system of claim 1 wherein the device comprises an ear-worn device, and wherein the PPG sensor comprised in the ear-worn device contacts a user's ear skin.
3. The assessment generation system of claim 1 wherein the processing circuit screens the data from the motion sensor by determining whether a user cadence is below a threshold to determine the integrity of the data from the PPG sensor.
4. The assessment generation system of claim 1 wherein the processing circuit is configured to generate the PPG waveform from an average pulse shape generated using the buffered PPG data.
5. The assessment generation system of claim 4 wherein the processing circuit is configured to generate the average pulse shape by averaging a spline representation of each of the buffered PPG data to generate an average spline representation.
6. The assessment generation system of claim 4 wherein the processing circuit is further configured to generate the PPG waveform by computing at least one of an integral of the average pulse shape and a derivative of the average pulse shape.
7. An ear-worn device configured to determine a blood pressure of a user, the ear-worn device comprising:
a motion sensor;
a photoplethysmography (PPG) sensor disposed proximate at least some skin of the user such that at least the PPG sensor contacts a user's skin; and
a processing circuit configured to:
screen motion data from the motion sensor to determine an integrity of PPG data from the PPG sensor;
process the PPG data from the PPG sensor, wherein at least some of the processed PPG data has integrity as determined responsive to the motion data, to generate a PPG waveform from the PPG data; and
process the PPG waveform using a machine learning model to generate the blood pressure of the user.
8. The ear-worn device of claim 7 wherein the processing circuit screens the data from the motion sensor by determining whether a user cadence is below a threshold to determine the integrity of the data from the PPG sensor.
9. The ear-worn device of claim 7 wherein the processing circuit is configured to generate the PPG waveform from an average pulse shape generated from the processed PPG data.
10. The ear-worn device of claim 9 wherein the processing circuit is configured to generate the average pulse shape by averaging a spline representation of each of the processed PPG data to generate an average spline representation.
11. The ear-worn device of claim 9 wherein the processing circuit is further configured to generate the PPG waveform by computing at least one of an integral of the average pulse shape and a derivative of the average pulse shape.
12. An assessment generation system for a wearable device, the assessment generation system comprising:
a motion sensor configured to generate a motion signal output;
a biometric sensor configured to generate a biometric signal output;
a signal extraction processing circuit configured to process information from the motion signal output and information from the biometric signal output to attenuate motion information from the biometric signal output to generate a cleaner biometric signal output with reduced noise; and
an assessment processing circuit, comprising a neural network, configured to determine a blood pressure characteristic for the user from the cleaner biometric signal.