System, software, and method for detection of sleep-disordered breathing using an electrocardiogram

• US 20060041201A1
• Filed: 08/23/2004
• Published: 02/23/2006
• Est. Priority Date: 08/23/2004
• Status: Active Grant

• Current Assignee: Board of Regents of the University of Texas System (University of Texas System)
• Original Assignee: Board of Regents of the University of Texas System (University of Texas System)
• Inventors: Behbehani, Khosrow, Burk, John R., Vijendra, Sridhar, Lucas, Edgar A.
• Granted Patent: US 7,343,198 B2
• Time in Patent Office: Days
• US Class Current: 600/521
• CPC Class Codes: A61B 5/352 Detecting R peaks, e.g. for synchronising diagnostic apparatus; Estimating R-R interval
• CPC Class Codes: A61B 5/4818 Sleep apnoea
• CPC Class Codes: A61B 5/7257 using Fourier transforms

Abstract

A system to form and store an electrocardiogram (ECG) signal derived from a cardiac electrical signal that includes an apparatus having a pair of the electrodes to connect to a patient to detect the cardiac electrical signal. A signal sampler samples the cardiac electrical signal to form the ECG signal. A data storage device stores the ECG signal. A computer communicates with the data storage device to retrieve the ECG signal for analysis by software stored in the memory of the computer. The software analyzes a morphology of the amplitude of a plurality of R-wave peaks contained within the ECG signal and/or analyzes a morphology of the area of a plurality of QRS complex pulses contained within the ECG signal.

Claims

1. A method for detecting sleep-disordered breathing, the method comprising:

(a) detecting a position and magnitude of a plurality of R-wave peak amplitudes within an electrocardiogram signal, variations in the magnitude of the plurality of R-wave peak amplitudes defining R-wave peak amplitude morphology for the plurality of R-wave peaks; and

(b) analyzing the R-wave peak amplitude morphology to detect sleep-disordered breathing.

2. A method according to claim 1, wherein the step of analyzing R-wave peak amplitude morphology includes the step of:

forming an R-wave peak sampled time series by interpolating between consecutive R-wave peak amplitude values of the electrocardiogram signal to form an R-wave peak value trend and by sampling the formed R-wave peak value trend.

3. A method according to claim 2, wherein the step of analyzing R-wave peak amplitude morphology includes the steps of:

determining a power spectral density of the R-wave peak sampled time series;

determining a power level in a preselected frequency band of the power spectral density; and

comparing the determined power level to a predetermined threshold level.

4. A method according to claim 1, wherein the step of analyzing R-wave peak amplitude morphology includes the steps of:

dividing the electrocardiogram signal into a first plurality of epochs;

forming an R-wave peak time series for each epoch in a subset of the first plurality of epochs defining a second plurality of epochs by interpolating between consecutive R-wave peak amplitude values within each epoch of the second plurality of epochs to form an R-wave peak value trend for each of the second plurality of epochs; and

determining a power spectral density of the R-wave peak time series for each of the second plurality of epochs.

5. A method according to claim 4, wherein the step of analyzing R-wave peak amplitude morphology includes the steps of:

determining a power level in a preselected frequency band of the power spectral density for each of the second plurality of epochs;

comparing the determined power level for each of the second plurality of epochs to a predetermined threshold level to determine if the predetermined threshold is exceeded; and

signaling the existence of sleep-disordered breathing when a preselected percentage of second plurality of epochs exceed the predetermined threshold level.

6. A method according to claim 5, wherein the preselected frequency band is approximately between 0 to 0.1 Hz.

7. A method of claim 1, further comprising the step of sampling a cardiac electrical signal from the patient at a sampling rate of at least 250 hertz to form the electrocardiogram signal.

8. A method according to claim 4, wherein the step of forming an R-wave peak time series includes the step of sampling uniformly the R-wave peak value trend for each of the second plurality of epochs.

9. A method according to claim 8, wherein the step of determining the power spectral density includes the step of enhancing estimation of a spectrum of the R-wave peak value time series by applying a window function.

10. A method according to claim 4, wherein the step of detecting a position and magnitude of a plurality of R-wave peak amplitudes includes the steps of:

filtering each of the first plurality of epochs to substantially remove baseline wander; and

rejecting any epoch of the first plurality of epochs having at least one of excessive baseline wander, amplifier saturation, and noise, the remaining epochs forming the second plurality of epochs.

11. A method for detecting sleep-disordered breathing, the method comprising:

(a) detecting a position of a plurality of QRS complex pulses within an electrocardiogram signal;

(b) determining a plurality of QRS complex pulse area values for a preselected time band coincident with a position of a corresponding plurality of R-wave peaks for the plurality of QRS complex pulses, variations in the area for each of the plurality of QRS complex pulse areas defining QRS complex pulse area morphology for the plurality of QRS complex pulses;

(c) forming a QRS complex pulse area value sampled time series by interpolating between consecutive QRS complex pulse area values of the electrocardiogram signal to form a QRS complex pulse area value trend and by evenly sampling the formed QRS complex pulse area value trend; and

(d) analyzing the QRS complex pulse area morphology for the plurality of QRS complex pulses by determining a power spectral density of the QRS complex pulse area value sampled time series and determining a power level in a preselected frequency band of the power spectral density.

12. A method according to claim 11, further comprising the step of:

comparing the determined power level to a predetermined threshold level to determine if the predetermined threshold level is exceeded.

13. A method according to claim 11, wherein the preselected frequency band is a first frequency band having a first power level, the method further comprising performing at least one of the following steps:

comparing the first power level to a first predetermined threshold level to determine if the first predetermined threshold level is exceeded, the preselected frequency band approximately between 0.0 to 0.1 Hz; and

comparing a second power level of a second frequency band to a second predetermined threshold level to determine if the second power level fails to exceed the second predetermined threshold level, the second preselected frequency band approximately between 0.2 to 0.3 Hz.

14. A method according to claim 11, further comprising the steps of:

dividing the electrocardiogram signal into a first plurality of epochs;

forming an QRS complex pulse area value sampled time series for each epoch in a subset of the first plurality of epochs defining a second plurality of epochs by interpolating between consecutive QRS complex pulse area values of each epoch in the second plurality of epochs to form an QRS complex pulse area value trend for each epoch in the second plurality of epochs and by sampling the formed QRS complex pulse area value trend for each of the second plurality of epochs; and

determining a power spectral density of the QRS complex pulse area value sampled time series for each of the second plurality of epochs.

15. A method according to claim 14, further comprising the steps of:

determining a power level in a preselected frequency band of the power spectral density for each of the second plurality of epochs;

comparing the determined power level for each of the second plurality of epochs to a predetermined threshold level to determine if the predetermined threshold level is exceeded; and

determining the existence of sleep-disordered breathing when a preselected percentage of second plurality of epochs exceed the predetermined threshold level.

16. A method according to claim 11, wherein the preselected time band is less than 100 milliseconds wide.

17. A method according to claim 11, further comprising the steps of:

determining a power level in a preselected frequency band of the power spectral density for each of the second plurality of epochs;

comparing the determined power level for each of the second plurality of epochs to a predetermined threshold level to determine if the predetermined threshold level is exceeded; and

determining the existence of sleep-disordered breathing when a preselected percentage of second plurality of epochs fails to exceed the predetermined threshold level.

18. A system to detect sleep-disordered breathing, the apparatus comprising:

an apparatus to form and store an electrocardiogram signal from a cardiac electrical signal, the apparatus including;

at least one electrocardiogram lead having a pair of the electrodes adapted to connect to a selected portion of a patient to detect the cardiac electrical signal, the signal having a plurality of QRS complex pulses,

a data sampler electrically connected to the at least one electrocardiogram lead to form the electrocardiogram signal by sampling the detected cardiac electrical signal,

a data storage device for storing the electrocardiogram signal;

a computer having a processor and a memory associated therewith, the processor positioned to communicate with the data storage device and the memory; and

R-wave peak amplitude morphology analyzing software stored in the memory of the computer for analyzing R-wave peak amplitude morphology of a plurality of R-waves peaks contained within the electrocardiogram signal, the R-wave peak amplitude morphology analyzing software including;

an R-wave peak amplitude determiner, adapted to receive the electrocardiogram signal, to determine a position and magnitude of the plurality of R-wave peaks within the electrocardiogram signal, variations in the magnitude of the plurality of R-wave peak amplitudes defining R-wave peak amplitude morphology for the plurality of R-wave peaks, and

an R-wave peak amplitude morphology analyzer, responsive to the R-wave peak amplitude determiner, to detect sleep-disordered breathing from the R-wave peak amplitude morphology.

19. A system according to claim 18, wherein the data sampler comprises an analog to a digital converter having a sampling rate of at least 250 Hz.

20. A system according to claim 18, wherein the R-wave peak amplitude morphology analyzer further comprises:

an R-wave peak value time series generator, responsive to the R-wave peak amplitude determiner, to interpolate between consecutive R-wave peak amplitude values of the electrocardiogram signal to thereby form an R-wave peak value trend defining an R-wave peak value time series;

a power spectral density determiner, responsive to the R-wave peak value time series generator, to calculate a power spectral density of the R-wave peak time series;

a power level determiner, responsive to the power spectral density determiner, to determine a power level in a preselected frequency band of the power spectral density; and

a power level comparator, responsive to the power level determiner and positioned to retrieve a predetermined power threshold level from memory, to compare the determined power level in the preselected frequency band to the predetermined threshold level.

21. A system according to claim 18,

wherein the R-wave peak amplitude determiner further comprises:

an electrocardiogram signal parser, adapted to receive the electrocardiogram signal, to parse the electrocardiogram signal into a first plurality of epochs, and

an R-wave peak detector, responsive to the electrocardiogram signal parser, to detect a plurality of R-wave peak positions and amplitudes values within a subset of the first plurality of epochs defining a second plurality of epochs; and

wherein the R-wave peak amplitude morphology analyzer further comprises;

an R-wave peak value time series generator, responsive to the R-wave peak amplitude determiner, to interpolate between consecutive R-wave peak amplitude values of each epoch of the second plurality of epochs to form an R-wave peak value trend defining a R-wave peak value time series for each epoch of the second plurality of epochs, and

a power spectral density determiner, responsive to the R-wave peak value time series generator, to determine a power spectral density for the R-wave peak value time series for each of the second plurality of epochs.

22. A system according to claim 21, wherein the R-wave peak amplitude morphology analyzer further comprises:

a power level determiner, responsive to the power spectral density determiner, to determine a power level for a preselected frequency band of the power spectral density for each of the second plurality of epochs;

a power level comparator, responsive to the power level determiner, to compare the determined power level for each of the second plurality of epochs to a predetermined threshold level to thereby separately determine if either a predetermined maximum threshold level is exceeded or a predetermined minimum threshold level has not been met for each of the second plurality of epochs; and

a statistical comparator, responsive to the power level comparator and positioned to retrieve a preselected sleep disorder percentage from memory, to compare a percentage of the second plurality of epochs having power levels exceeding the predetermined threshold level to the preselected sleep disorder percentage, to thereby determine a probable existence of sleep-disordered breathing when the percentage of the second plurality of epochs having power levels either exceeding the predetermined maximum threshold level or failing to meet the minimum threshold level exceeds the preselected sleep disorder percentage.

23. A system according to claim 21, wherein the power spectral density determiner includes:

an algorithm to subdivide each epoch of the second plurality of epochs into a plurality of overlapping sections;

an algorithm to determine the power spectral density for each of the plurality of sections within each epoch; and

an algorithm to average the resultant power spectral density for each section within each epoch to form the power spectral density for each epoch.

24. A system according to claim 23, wherein the power spectral density determiner further includes a window algorithm to enhance estimation of a spectrum of the R-wave peak value time series.

25. A system according to claim 21, wherein R-wave peak amplitude morphology analyzer includes:

a baseline filter, responsive to the electrocardiogram signal parser, to filter each of the first plurality of epochs to thereby substantially remove baseline wander; and

a screener, responsive to the baseline filter, to reject any epoch of the first plurality of epochs having excessive baseline wander.

26. A system according to claim 22, wherein the preselected frequency band is a first preselected frequency band the power spectral density of each epoch approximately between 0 to 0.1 Hz and assigned the predetermined threshold level, and wherein the power spectral density of the each epoch includes a second preselected frequency band approximately between 0.2 to 0.3 Hz and assigned the predetermined minimum threshold level.

27. A system according to claim 18, further comprising QRS complex pulse area morphology analyzing software stored in the memory of the computer for analyzing a morphology of an area value for a plurality of QRS complex pulses in an electrocardiogram signal, the QRS complex pulse area morphology analyzing software including:

a QRS complex pulse area value determiner to determine the QRS complex pulse area values of the plurality of QRS complex pulses for a plurality of electrocardiogram signal epochs, each QRS complex pulse area value determined for a preselected time band less than 100 milliseconds wide and coincident with a position of corresponding R-wave peaks in the electrocardiogram signal; and

a QRS complex pulse area value morphology analyzer, responsive to the QRS complex pulse area value determiner, to detect sleep-disordered breathing from morphology of the area value for the plurality of QRS complex pulses in the electrocardiogram signal, the analyzer including;

a QRS complex pulse area value time series generator, responsive to the QRS complex pulse area determiner, to interpolate between consecutive QRS complex pulse area values of each of the second plurality of epochs to form an QRS complex pulse area value trend and to uniformly sample the QRS complex pulse area value trend to form an evenly sampled time series defining a QRS complex pulse area value time series for each of the plurality of epochs, and

a power spectral density determiner, responsive to the QRS complex pulse area value time series generator, to determine a power spectral density for the QRS complex pulse area value time series for each of the plurality of epochs.

28. R-wave peak amplitude morphology analyzing software stored on a storage media for analyzing a morphology of a plurality of R-waves contained within the electrocardiogram signal, the R-wave peak amplitude morphology analyzing software including:

an R-wave peak amplitude determiner, adapted to receive the electrocardiogram signal, to determine a position and magnitude of the plurality of R-wave peaks within the electrocardiogram signal, variations in the magnitude of the plurality of R-wave peak amplitudes defining R-wave peak amplitude morphology for the plurality of R-wave peaks; and

an R-wave peak amplitude morphology analyzer, responsive to the R-wave peak amplitude determiner, to detect sleep-disordered breathing from the R-wave peak amplitude morphology.

29. Software as defined in claim 28, wherein the R-wave peak amplitude morphology analyzer further comprises:

an R-wave peak value time series generator, responsive to the R-wave peak amplitude determiner, to interpolate between consecutive R-wave peak amplitude values of the electrocardiogram signal to thereby form an R-wave peak value trend.

30. Software as defined in claim 29, wherein the R-wave peak amplitude morphology analyzer further comprises:

a power spectral density determiner, responsive to the R-wave peak value time series generator, to calculate a power spectral density of the R-wave peak time series;

a power level determiner, responsive to the power spectral density determiner, to determine a power level in a preselected frequency band of the power spectral density; and

a power level comparator, responsive to the power level determiner and positioned to retrieve a predetermined power threshold level from memory, to compare the determined power level in the preselected frequency band to the predetermined threshold level.

31. Software as defined in claim 28, wherein the R-wave peak amplitude determiner further comprises an electrocardiogram signal parser, adapted to receive the electrocardiogram signal, to parse the electrocardiogram signal into a first plurality of epochs, and wherein the R-wave peak amplitude morphology analyzer further comprises:

an R-wave peak value time series generator, responsive to the electrocardiogram signal parser of the R-wave peak amplitude determiner, to interpolate between consecutive R-wave peak amplitude values of each epoch in a subset of the first plurality of epochs defining a second plurality of epochs to form an R-wave peak value trend defining a R-wave peak value time series for each of the second plurality of epochs, and

a power spectral density determiner, responsive to the R-wave peak value time series generator, to determine a power spectral density for the R-wave peak value time series for each of the second plurality of epochs.

32. Software as defined in claim 31, wherein the R-wave peak amplitude morphology analyzer further comprises:

a power level determiner, responsive to the power spectral density determiner, to determine a power level for a preselected frequency band of the power spectral density for each of the second plurality of epochs; and

a power level comparator, responsive to the power level determiner, to compare the determined power level for each of the second plurality of epochs to a predetermined threshold level to thereby separately determine if a predetermined threshold is exceeded for each of the second plurality of epochs.

33. Software as defined in claim 32, wherein the R-wave peak amplitude morphology analyzer further comprises:

a statistical comparator, responsive to the power level comparator and positioned to retrieve a preselected sleep disorder percentage from memory, to compare a percentage of the second plurality of epochs having power levels exceeding the predetermined threshold level to the preselected sleep disorder percentage to thereby determine a probable existence of sleep-disordered breathing when the percentage of the second plurality of epochs having power levels exceeding the predetermined threshold level exceeds the preselected sleep disorder percentage.

34. Software as defined in claim 31, wherein the R-wave peak value time series generator samples the R-wave peak value trend for each of the second plurality of epochs to form an evenly sampled R-wave peak value time series for each of the second plurality of epochs.

35. Software as defined in claim 34, wherein the power spectral density determiner further includes a window algorithm to enhance estimation of a spectrum of the R-wave peak value time series.

36. Software as defined in claim 31, wherein R-wave peak amplitude morphology analyzer includes:

a baseline filter, responsive to the electrocardiogram signal parser, to filter each of the first plurality of epochs to thereby substantially remove baseline wander; and

a screener, responsive to the epoch baseline filter, to reject any epoch of the first plurality of epochs having excessive baseline wander.

37. A computer readable medium that is readable by a computer facilitating detecting sleep disordered breathing in a patient, the computer readable medium comprising a set of instructions that, when executed by the computer, cause the computer to perform the following operations:

detect a position and magnitude of a plurality of R-wave peak amplitudes within at least a portion of an electrocardiogram signal;

form a R-wave peak sampled time series by forming an R-wave peak value trend and sampling the formed trend;

determine a power spectral density of the R-wave peak sampled time series;

determine a power level in a preselected frequency band of the power spectral density; and

compare the determined power level to a predetermined threshold level.

38. A computer readable medium that is readable by a computer facilitating detecting sleep disordered breathing in a patient, the computer readable medium comprising a set of instructions that, when executed by the computer, cause the computer to perform the following operations:

divide an electrocardiogram signal into a first plurality of epochs;

form an R-wave peak time series for each epoch in a subset of the first plurality of epochs defining a second plurality of epochs by forming an R-wave peak value trend and sampling the formed trend;

determine a power spectral density of the R-wave peak time series for each of the second plurality of epochs;

determine a power level in a preselected frequency band of the power spectral density for each of the second plurality of epochs;

compare the determined power level for each of the second plurality of epochs to a predetermined threshold level to determine if the predetermined threshold is exceeded; and

signal the existence of sleep-disordered breathing when a preselected percentage of second plurality of epochs exceed the predetermined threshold level.