# US 11,998,347 B2 · Sleep apnea detection system and method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 11,998,347 B2
- Document ID: US11998347B2
- Status: Granted
- Kind: B2
- Application number: 16/977,595
- Filed: 2019-03-06
- Published: 2020-12-31
- Granted: 2024-06-04
- Earliest claimed: 2018-03-07
- Assignees: Koninklijke Philips N.V.
- Applicants: Koninklijke Philips N.V.
- Inventors: Koen Theo Johan De Groot; Pedro Miguel Fonseca; Jenny Margarito
- Web page: https://patentgrep.com/patent/US11998347B2

## CPC

- A61B 5/4818: {Sleep apnoea}
- A61B 5/02405: {Determining heart rate variability}
- A61B 5/02416: {using photoplethysmograph signals, e.g. generated by infrared radiation (A61B5/14552 takes precedence)}
- A61B 5/4812: {Detecting sleep stages or cycles}
- A61B 5/681: {Wristwatch-type devices}
- A61B 5/7264: {Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems}
- A61B 5/7267: {involving training the classification device}
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

A method and system for detecting sleep apnea involves determining the sleep stage, and detecting an apnea event based on a physiological sensor signal using selection of a detection algorithm which is dependent on the determined sleep stage. By taking account of the sleep stage when performing an automated apnea detection process, the accuracy of the apnea detection is improved.

## Claims

### Claim 1

A system for detecting sleep apnea, comprising:
a physiological sensor for generating a sensor signal from a user for use in detecting sleep apnea in the user; and
a processor coupled to the physiological sensor, the processor being structured and configured to include:
a sleep stage detection unit including a sleep stage classifier structured and configured to determine a sleep stage of the user based on heart rate variability information derived from the sensor signal;
a plurality of pre-trained apnea classification models, each pre-trained sleep apnea classification model being associated with a particular one of a plurality of predetermined sleep stages and being structured and configured to automatically detect an apnea event based on the heart rate variability information, the plurality of predetermined sleep stages including at least two different non-awake sleep states;
an apnea classification model selection unit structured and configured to select one of the pre-trained apnea classification models based on the determined sleep stage of the user; and
an apnea detection unit structured and configured to determine whether apnea has been detected in the user using the selected one of the pre-trained apnea classification models and the heart rate variability information and to generate an output indicative of whether apnea has been detected.

### Claim 2 (depends on 1)

The system as claimed in claim 1, wherein the physiological sensor comprises a PPG sensor and/or an ECG sensor.

### Claim 3 (depends on 1)

The system as claimed in claim 1, wherein the processor is adapted to: extract heart beat timings from the sensor signal; derive an interbeat interval time series; and extract the heart rate variability information from the interbeat interval time series.

### Claim 4 (depends on 3)

The system as claimed in claim 3, wherein the sleep stage classifier is structured and configured to: sample the sensor signal as a sequence of time frames; extract first features comprising the heart rate variability information from the sensor signal from each time frame; and determine the sleep stage for each time frame from the first features.

### Claim 5 (depends on 1)

The system as claimed in claim 1, wherein the heart rate variability information is derived using a model which depends on a previously identified sleep stage.

### Claim 6

A method for detecting sleep apnea, comprising:
generating a sensor signal from a user for use in detecting sleep apnea using a physiological sensor;
determining a sleep stage of a subject during sleep based on heart rate variability information derived from the sensor signal using a sleep stage classifier of a sleep stage detection unit implemented as part of a processor, wherein the processor includes a plurality of pre-trained apnea classification models, each pre-trained sleep apnea classification model being associated with a particular one of a plurality of predetermined sleep stages and being structured and configured to automatically detect an apnea event based on the heart rate variability information, the plurality of predetermined sleep stages including at least two different non-awake sleep states;
selecting one of the pre-trained apnea classification models based on the determined sleep stage of the user using an apnea classification model selection unit implemented in the processor;
determining, in an apnea detection unit implemented in the processor, whether apnea has been detected in the user using the selected one of the pre-trained apnea classification models and the heart rate variability information; and
generating an output from the processor indicative of whether apnea has been detected.

### Claim 7 (depends on 6)

The method as claimed in claim 6, comprising: extracting heart beat timings from the sensor signal; deriving an interbeat interval timeseries; and extracting the heart rate variability information from the interbeat interval time series.

### Claim 8 (depends on 7)

The method as claimed in claim 7, comprising: sampling the sensor signal as a sequence of time frames; extracting first features comprising the heart rate variability information from of the sensor signal from each time frame; and determining a sleep stage for each time frame from the first features.

### Claim 9 (depends on 6)

A computer program product, comprising a non-transitory computer usable medium having computer program code embodied therein which is adapted to implement the method for detecting sleep apnea as recited in claim 6.