Determining Sleep Stages and Sleep Events Using Sensor Data

• US 20170055898A1
• Filed: 08/26/2016
• Published: 03/02/2017
• Est. Priority Date: 08/28/2015
• Status: Active Grant

• Current Assignee: Awarables, Inc.
• Original Assignee: Awarables, Inc.
• Inventors: Upender, Raghu, Giles, Chris, Bandyopadhyay, Amrit, Blankenship, Gilmer, Upender, Madhvi
• Granted Patent: US 10,321,871 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: A61B 2503/12 Healthy persons not otherwise provided for, e.g. subjects of a marketing survey
• CPC Class Codes: A61B 2560/0242 adapted to measure environmental factors, e.g. temperature, pollution
• CPC Class Codes: A61B 2562/0219 Inertial sensors, e.g. accelerometers, gyroscopes, tilt switches
• CPC Class Codes: A61B 2562/06 Arrangements of multiple sensors of different types
• CPC Class Codes: A61B 5/0022 Monitoring a patient using a global network, e.g. telephone networks, internet
• CPC Class Codes: A61B 5/0205 Simultaneously evaluating both cardiovascular conditions and different types of body conditions, e.g. heart and respiratory condition
• CPC Class Codes: A61B 5/02405 Determining heart rate variability
• CPC Class Codes: A61B 5/02416 using photoplethysmograph signals, e.g. generated by infrared radiation A61B5/14552 takes precedence
• CPC Class Codes: A61B 5/02438 with portable devices, e.g. worn by the patient
• CPC Class Codes: A61B 5/0816 Measuring devices for examining respiratory frequency measuring frequency of electric signals G01R23/00
• CPC Class Codes: A61B 5/0826 Detecting or evaluating apnoea events A61B5/4818 takes precedence
• CPC Class Codes: A61B 5/11 Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb for measuring pulse A61B5/02 ;A61B5/1038 takes precedence; motion detection to correct for motion artifacts in physiological signals A61B5/721
• CPC Class Codes: A61B 5/1118 Determining activity level
• CPC Class Codes: A61B 5/30 Input circuits therefor
• CPC Class Codes: A61B 5/303 Patient cord assembly, e.g. cable harness
• CPC Class Codes: A61B 5/316 Modalities, i.e. specific diagnostic methods
• CPC Class Codes: A61B 5/318 Heart-related electrical modalities, e.g. electrocardiography [ECG]
• CPC Class Codes: A61B 5/352 Detecting R peaks, e.g. for synchronising diagnostic apparatus; Estimating R-R interval
• CPC Class Codes: A61B 5/369 Electroencephalography [EEG] devices for psychotechnics A61B5/16
• CPC Class Codes: A61B 5/4064 Evaluating the brain A61B5/031, A61B5/369, A61B5/14553 take precedence
• CPC Class Codes: A61B 5/4806 : Sleep evaluation A61B5/4821 takes precedence; devices for inducing sleep A61M21/02
• CPC Class Codes: A61B 5/4812 : Detecting sleep stages or cycles
• CPC Class Codes: A61B 5/4818 : Sleep apnoea
• CPC Class Codes: A61B 5/486 : Bio-feedback A61B5/375 takes precedence
• CPC Class Codes: A61B 5/6804 : Garments; Clothes
• CPC Class Codes: A61B 5/681 : Wristwatch-type devices
• CPC Class Codes: A61B 5/6823 : Trunk, e.g., chest, back, abdomen, hip
• CPC Class Codes: A61B 5/6898 : Portable consumer electronic devices, e.g. music players, telephones, tablet computers
• CPC Class Codes: A61B 5/7246 : using correlation, e.g. template matching or determination of similarity
• CPC Class Codes: A61B 5/7257 : using Fourier transforms
• CPC Class Codes: A61B 5/7267 : involving training the classification device
• CPC Class Codes: A61B 5/7278 : Artificial waveform generation or derivation, e.g. synthesising signals from measured signals A61B5/327 takes precedence
• CPC Class Codes: A61B 5/742 : using visual displays A61B5/339 takes precedence
• CPC Class Codes: G16H 50/20 : for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for determining sleep stages and sleep events using sensor data. In some implementations, sensor data is obtained over a time period while a person is sleeping. The time period is divided into a series of intervals. Heart rate and changes in the heart rate are analyzed over the intervals. Based on the analysis of the heart rate changes, sleep stage labels are assigned to different portions of the time period. An indication of the assigned sleep stage labels is provided.

Claims

1. A method performed by one or more computing devices, the method comprising:

obtaining, by the one or more computing devices, sensor data generated by one or more sensors over a time period while a person is sleeping;

dividing, by the one or more computing devices, the time period into a series of intervals;

analyzing, by the one or more computing devices, heart rate and the changes in the heart rate of the person indicated by the sensor data over the intervals;

based on the analysis of heart rate the heart rate changes, assigning, by the one or more computing devices, sleep stage labels to different portions of the time period; and

providing, by the one or more computing devices, an indication of the assigned sleep stage labels.

2. The method of claim 1, wherein obtaining the sensor data, dividing the time period into a series of intervals, and analyzing the heart rate and the changes in the heart rate are performed by a wearable, battery-operated device, wherein the wearable device includes sensors that detect the sensor data.

3. The method of claim 2, further comprising providing, by the wearable device, results of analyzing the heart rate and the changes in the heart rate to a second device for transmission to a server system;

and

wherein assigning the sleep stage labels and providing the indication of the assigned sleep stage labels are performed by the server system.

4. The method of claim 1, wherein the sensor data comprises movement data from one or more motion sensors;

and

wherein the method comprises determining correlations of the movement data with sensor data indicating heart beats of the person, wherein the sleep stage labels are assigned based at least in part on the correlations between the movement data and the sensor data indicating the heart beats.

5. The method of claim 1, wherein assigning a sleep stage label to a particular portion of the time period comprises:

obtaining likelihood scores from multiple different sleep stage analysis functions; and

performing a sleep stage classification decision for the particular portion of the time period based on a combination of the likelihood scores from the multiple different sleep stage analysis functions.

6. The method of claim 1, wherein the sleep stage labels include labels corresponding to stages for wake, rapid eye movement (REM) sleep, light sleep, and deep or slow-wave sleep.

7. The method of claim 1, wherein assigning the sleep stage labels comprises:

determining, for each of the intervals, (i) a likelihood that the interval corresponds to REM sleep, and (ii) an indication whether the interval is classified as REM sleep;

determining, for each of the intervals, (i) a likelihood that the interval corresponds to light sleep, and (ii) an indication whether the interval is classified as light sleep;

determining, for each of the intervals, (i) a likelihood that the interval corresponds to slow-wave sleep, and (ii) an indication whether the interval is classified as slow-wave sleep;

or

determining, for each of the intervals, (i) a likelihood that the interval corresponds to a wake stage, and (ii) an indication whether the interval is classified as a wake stage.

8. The method of claim 1, wherein assigning the sleep stage labels comprises determining, for each of the intervals, (i) a likelihood that the interval corresponds to the person being awake or asleep, and (ii) an indication whether the person is classified as being awake or asleep.

9. The method of claim 1, wherein the sensor data comprises at least one selected from a group consisting of:

(i) EKG signal data from EKG sensor, and (ii) heartbeat data from a heartbeat sensor.

10. The method of claim 1, wherein dividing the time period into a series of intervals comprises diving the time period into adjacent periods each having a same duration.

11. The method of claim 10, wherein dividing the time period into a series of intervals comprises overlapping sliding windows having the same duration, with a sliding window being centered at each sample of heart rate data.

12. The method of claim 1, wherein analyzing the changes in the heart rate of the person comprises determining heart rate variability characteristic scores for different portions of the sleep session, and wherein the sleep stage labels are assigned based at least in part on the heart rate variability scores.

13. The method of claim 1, wherein analyzing the changes in the heart rate of the person comprises determining measures of randomness of heartbeat data for different sliding windows of the time period, wherein the sleep stage labels are assigned based at least in part on the heart rate variability scores.

14. The method of claim 13, wherein the measure of randomness of heartbeat data is computed by determining a measure of approximate entropy based on the heartbeat data.

15. The method of claim 1, wherein analyzing the heart rate of the person comprises evaluating the value of an EKG signal in each interval with respect to the values in a neighborhood of the interval, wherein the neighborhood for each interval is a time window that extends from a first time threshold that precedes the interval to a second time threshold that follows the interval.

16. The method of claim 1, wherein the obtained sensor data corresponds to a sleep session of the person;

and

wherein analyzing the heart rate of the person comprises evaluating an absolute value of the heart rate signal with respect to the rest of the heart rate values for the sleep session and predetermined thresholds.

17. The method of claim 1, wherein analyzing the changes in the heart rate of the person comprises:

generating a heart-rate variability (HRV) signal; and

performing a frequency analysis of the HRV signal; and

examining a ratio of low frequency components of the HRV signal to high frequency components of the HRV signal.

18. The method of claim 1, wherein assigning a sleep stage label to a particular portion of the time period comprises:

obtaining likelihood scores from multiple different sleep stage analysis functions; and

performing a sleep stage classification decision for the particular portion of the time period based on a combination of the likelihood scores from the multiple different sleep stage analysis functions.

19. The method of claim 1, wherein assigning a sleep stage label to a particular portion of the time period comprises:

providing multiple signals that are separately correlated to sleep stages as input to a sleep stage classifier; and

obtaining a single sleep stage label for the particular portion of the time period based on output of the sleep stage classifier.

20. The method of claim 1, further comprising:

obtaining data sets indicating signals or function outputs, the data sets being labeled with sleep stage labels corresponding to the data sets; and

training the sleep stage classifier based on the data sets to produce output indicating likelihoods that input data corresponds to a sleep stage from among a set of sleep stages.

21. The method of claim 20, further comprising:

determining, for each of the signals or function outputs, a signal range histogram for the signal or function output; and

using the signal range histograms to train the sleep stage classifier.

22. The method of claim 1, wherein analyzing the changes in heart rate comprises detecting heartbeats indicated by the sensor data.

23. The method of claim 22, wherein detecting the heartbeats comprises:

detecting peaks in an EKG signal;

determining a derivative signal from of the EKG signal by determining, for each sample of the EKG signal, a difference of between a current sample and the immediately previous sample;

identifying R waves based on applying one or more thresholds to the derivative signal; and

detecting the heartbeats based on the identified R waves.

24. The method of claim 11, wherein detecting the heartbeats comprises:

identifying local maxima and local minima in an EKG signal;

computing ranges between the maximum value of the EKG signal and minimum value of the EKG signal within each of multiple windows of the EKG signal; and

assigning, to each of the multiple windows, a probability of the window representing a beat.

25. A system comprising:

one or more computing devices;

one or more computer-readable media storing instructions that, when executed by the one or more computing devices, cause the one or more computing devices to perform operations comprising;

obtaining, by the one or more computing devices, sensor data generated by one or more sensors over a time period while a person is sleeping;

dividing, by the one or more computing devices, the time period into a series of intervals;

analyzing, by the one or more computing devices, heart rate and the changes in the heart rate of the person indicated by the sensor data over the intervals;

based on the analysis of the heart rate changes, assigning, by the one or more computing devices, sleep stage labels to different portions of the time period; and

providing, by the one or more computing devices, an indication of the assigned sleep stage labels.