Detection of periodic breathing

Patent metadata

Field	Value
Publication number	US-10159421-B2
Application number	US-201615079339-A
Country	US
Kind code	B2
Filing date	Mar 24, 2016
Priority date	Mar 30, 2015
Publication date	Dec 25, 2018
Grant date	Dec 25, 2018

Abstract

Methods and apparatus perform periodic breathing detection, such as Cheyne-Stokes respiration detection. The detection may be performed by one or more processors, such as by analysis of data from one or more sensors. In some cases, the detection may be based on an electrocardiogram (ECG) signal, such as from ECG electrodes and/or an accelerometer signal, such as from an accelerometer. An occurrence of periodic breathing may be detected based on features derived from the signal(s). For example, detection may be based on deriving a respiration signal from the sensed signal(s) and/or analysis of RR interval times or relative QRS amplitude values, which may be evaluated on a segment-by-segment basis. The detection may provide monitoring and reporting of the occurrence of periodic breathing by a monitoring device and/or provide a basis for controlling changes to a provided respiratory treatment or therapy, such as by a respiratory pressure therapy device.

First claim

The invention claimed is:

1. An apparatus for detecting periodic breathing in a patient, the apparatus comprising:
a processor configured to:
receive an electrocardiogram (ECG) signal of the patient;
derive a feature from the ECG signal; and
analyze the feature derived from the ECG signal to determine an occurrence of periodic breathing,

wherein the processor is further configured to:
receive an accelerometer signal indicative of the patient's position,
derive a feature from the accelerometer signal, and
analyze the feature derived from the accelerometer signal to determine an occurrence of periodic breathing,
wherein the feature derived from the accelerometer signal is a power spectral density of a demodulated envelope signal of the accelerometer signal.

2. The apparatus of claim 1 , further comprising a memory for storing the ECG signal.

3. The apparatus of claim 1 , further comprising a sensor to measure the ECG signal from the patient.

4. The apparatus of claim 3 , wherein the sensor is a Holter monitor.

5. The apparatus of claim 3 , wherein the sensor is a 12-lead ECG.

6. The apparatus of claim 3 , wherein the sensor is a patch type ECG.

7. The apparatus of claim 1 , wherein the processor is configured to perform a time-domain analysis of the ECG signal.

8. The apparatus of claim 1 , wherein the processor is configured to perform a frequency-domain analysis of the ECG signal.

9. The apparatus of claim 1 , wherein the processor is configured to divide the ECG signal into a plurality of time segments of equal time length.

10. The apparatus of claim 9 , wherein the processor is configured to determine whether each time segment of the plurality of time segments exhibits a characteristic of periodic breathing.

11. The apparatus of claim 1 , wherein the processor determines a likelihood of the patient having periodic breathing.

12. The apparatus of claim 1 , wherein the processor is configured to derive a respiratory signal from the ECG signal.

13. The apparatus of claim 12 , wherein the processor is configured to analyze an envelope of the derived respiratory signal.

14. The apparatus of claim 1 , wherein the feature derived from the ECG signal is a power spectral density of RR intervals in the ECG signal.

15. The apparatus of claim 1 , wherein the feature derived from the ECG signal is a power spectral density of ECG-derived respiration (EDR) numbers.

16. The apparatus of claim 15 , wherein an EDR number is a magnitude of a QRS peak in the ECG signal.

17. The apparatus of claim 15 , wherein an EDR number is an integral of an area around a QRS peak in the ECG signal.

18. The apparatus of claim 1 , wherein the processor determines the occurrence of periodic breathing by comparing the feature in a respiration frequency range to a predetermined threshold.

19. The apparatus of claim 1 , wherein the processor is configured to perform baseline correction on the ECG signal.

20. The apparatus of claim 1 , further comprising a sensor to measure the accelerometer signal.

21. The apparatus of claim 1 , further comprising a sensing device configured to measure the accelerometer signal and the ECG signal.

22. The apparatus of claim 21 , wherein the sensing device is a patch type ECG.

23. The apparatus of claim 1 , wherein the processor is further configured to derive an additional feature from the accelerometer signal, and wherein the additional feature derived from the accelerometer signal is a respiratory effort feature.

24. The apparatus of claim 1 , wherein the processor is configured to remove a movement artefact from the accelerometer signal.

25. The apparatus of claim 1 , wherein the periodic breathing is Cheyne-Stokes respiration.

26. The apparatus of claim 1 , wherein the processor is configured to combine features from the accelerometer signal and the ECG signal, in order to determine an occurrence of periodic breathing.

27. The apparatus of claim 26 , wherein the combined features comprise RR-interval, EDR and Respiratory Effort extracted features.

28. The apparatus of claim 1 wherein to analyze the feature derived from the ECG signal, the processor determines power spectrum of RR interval times or relative QRS amplitude values on a segment-by-segment basis.

29. The apparatus of claim 28 wherein the processor is configured to integrate the power spectrum in a predetermined range to output a CSR band power value.

30. The apparatus of claim 1 wherein the processor is configured to compare CSR band power values to a predetermined threshold to detect significant CSR band power values.

31. The apparatus of claim 1 wherein the processor is configured to count significant CSR band power values.

32. The apparatus of claim 31 wherein the processor is configured to present a ratio of the count of significant CSR band power values to a total number of time segments.

33. The apparatus of claim 1 wherein the processor is configured to determine an average CSR frequency or average cycle length from time segments selected according to significant CSR band power values.

34. The apparatus of claim 1 wherein the analysis of the feature derived from the accelerometer signal and the analysis of the feature derived from the ECG signal comprises classifying periodic breathing, in a classifier, with the derived feature from the accelerometer signal and the derived feature from the ECG signal.

35. The apparatus of claim 1 wherein the analysis of the feature derived from the accelerometer signal and the analysis of the feature derived from the ECG signal comprises classifying Cheyne-Stokes respiration, in a classifier, with the derived feature from the accelerometer signal and the derived feature from the ECG signal.

36. A method for detecting periodic breathing in a patient, the method comprising:
receiving, by a processor, an electrocardiogram (ECG) signal of the patient;
deriving, by the processor, a feature from the ECG signal;
analyzing, by the processor, the feature derived from the ECG signal to determine an occurrence of periodic breathing,
receiving an accelerometer signal indicative of the patient's position,
deriving a feature from the accelerometer signal, and
analyzing the feature derived from the accelerometer signal to determine an occurrence of periodic breathing,
wherein the feature derived from the accelerometer signal is a power spectral density of a demodulated envelope signal of the accelerometer signal.

37. The method of claim 36 , further comprising retrieving the ECG signal from a memory.

38. The method of claim 36 , wherein the ECG signal is provided by a sensor.

39. The method of claim 36 , further comprising performing a time-domain analysis of the ECG signal.

40. The method of claim 36 , further comprising performing a frequency-domain analysis of the ECG signal.

41. The method of claim 36 , further comprising dividing the ECG signal into a plurality of time segments of equal time length.

42. The method of claim 41 , further comprising determining whether each time segment of the plurality of time segments exhibits a characteristic of periodic bre

Assignees

• Resmed Sensor Tech Ltd

Inventors

• Heneghan Conor

Classifications

• G16H50/20

  for computer-aided diagnosis, e.g. based on medical expert systems · CPC title

• A61B5/7278

  Artificial waveform generation or derivation, e.g. synthesizing signals from measured signals · CPC title

• A61B5/4818

  Sleep apnoea · CPC title

• A61B5/113

  occurring during breathing · CPC title

• A61B5/7264 Primary

  Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems · CPC title