System and method for determining spectral boundaries for sleep stage classification

Patent metadata

Field	Value
Publication number	US-10702207-B2
Application number	US-201515533084-A
Country	US
Kind code	B2
Filing date	Dec 4, 2015
Priority date	Dec 11, 2014
Publication date	Jul 7, 2020
Grant date	Jul 7, 2020

Abstract

The present disclosure pertains to a system ( 10 ) configured to determine spectral boundaries ( 216, 218 ) for sleep stage classification in a subject ( 12 ). The spectral boundaries may be customized and used for sleep stage classification in an individual subject. Spectral boundaries determined by the system that are customized for the subject may facilitate sleep stage classification with higher accuracy relative to classifications made based on static, fixed spectral boundaries that are not unique to the subject. In some implementations, the system comprises one or more of a sensor ( 16 ), a processor ( 20 ), electronic storage ( 22 ), a user interface ( 24 ), and/or other components.

First claim

What is claimed is:

1. A system configured to determine subject specific spectral boundaries for sleep stage classification in a subject, the system comprising:
one or more sensors configured to generate output signals that convey information related to a respiratory wave amplitude metric for a sleep session of the subject; and
one or more physical computer processors configured by computer readable instructions to:
transform the information conveyed by the output signals in a first set of individual epochs of time into a frequency domain;
determine individual frequencies of respiratory wave amplitude metric peaks within the first set of individual epochs of time;
determine an aggregated frequency of the respiratory wave amplitude metric peaks by aggregating the determined individual frequencies of the respiratory wave amplitude metric peaks;
determine subject specific upper spectral boundaries and lower spectral boundaries for sleep stage classification for the subject based on the aggregated frequency, wherein the upper spectral boundaries are a function of the aggregated frequency and upper coefficients, and wherein the lower spectral boundaries are a function of the aggregated frequency and lower coefficients; and
determine sleep stages of the subject during a second set of individual epochs of time in a subsequent sleep session as a function of the aggregated frequency of the respiratory wave amplitude metric peaks using the determined subject specific upper and lower boundaries.

2. The system of claim 1 , wherein the one or more sensors and the one or more physical computer processors are configured such that the respiratory wave amplitude metric is a power spectral density.

3. The system of claim 2 , wherein the one or more physical computer processors are configured such that determining the aggregated frequency of the respiratory wave amplitude metric peaks comprises averaging frequencies of power spectral density peaks from the first set of individual epochs of time.

4. The system of claim 3 , wherein the one or more physical computer processors are configured such that an average frequency of the power spectral density peaks from individual thirty second epochs of time during the sleep session is a mean respiratory frequency of the subject.

5. The system of claim 4 , wherein the one or more physical computer processors are configured such that the subject specific upper spectral boundaries are determined based on the mean respiratory frequency using linear regression and upper regression coefficients, and the lower spectral boundaries are determined based on the mean respiratory frequency using linear regression and lower regression coefficients.

6. A method to determine subject specific spectral boundaries for sleep stage classification in a subject with a determination system, the determination system comprising one or more sensors and one or more physical computer processors, the method comprising:
generating, with the one or more sensors, output signals that convey information related to a respiratory wave amplitude metric for a sleep session of the subject;
transforming, with the one or more physical computer processors, the information conveyed by the output signals in a first set of individual epochs of time into a frequency domain;
determining, with the one or more physical computer processors, individual frequencies of respiratory wave amplitude metric peaks within the first set of individual epochs of time;
determining, with the one or more physical computer processors, an aggregated frequency of the respiratory wave amplitude metric peaks by aggregating the individual frequencies of the respiratory wave amplitude metric peaks within the first set of individual epochs of time;
determining, with the one or more physical computer processors, subject specific upper spectral boundaries and lower spectral boundaries for sleep stage classification for the subject based on the aggregated frequency, wherein the upper spectral boundaries are a function of the aggregated frequency and upper coefficients, and wherein the lower spectral boundaries are a function of the aggregated frequency and lower coefficients; and
determining, with the one or more physical computer processors, sleep stages of the subject during a second set of individual epochs of time in a subsequent sleep session as a function of the aggregated frequency of the respiratory wave amplitude metric peaks using the determined subject specific upper and lower spectral boundaries.

7. The method of claim 6 , wherein the respiratory wave amplitude metric is a power spectral density.

8. The method of claim 7 , wherein determining the aggregated frequency of the respiratory wave amplitude metric peaks comprises averaging frequencies of power spectral density peaks from the first set of individual epochs of time.

9. The method of claim 8 , wherein an average frequency of the power spectral density peaks from individual thirty second epochs of time during the sleep session is a mean respiratory frequency of the subject.

10. The method of claim 9 , wherein the subject specific upper spectral boundaries are determined based on the mean respiratory frequency using linear regression and upper regression coefficients, and the lower spectral boundaries are determined based on the mean respiratory frequency using linear regression and lower regression coefficients.

Assignees

• Koninklijke Philips Nv

Inventors

• Long Xi
• Haakma Reinder
• Ferreira Dos Santos Da Fonseca Pedro Miguel
• Aarts Ronaldus Maria

Classifications

• A61B5/1135

  by monitoring thoracic expansion · CPC title

• A61B5/02416

  using photoplethysmograph signals, e.g. generated by infrared radiation (A61B5/14552 takes precedence) · CPC title

• A61B5/1118

  Determining activity level · CPC title

• A61B5/681

  Wristwatch-type devices · CPC title

• A61B5/7257

  using Fourier transforms · CPC title