Apparatus for detecting sleep apnea using electrocardiogram signals

• US 20030055348A1
• Filed: 09/14/2001
• Published: 03/20/2003
• Est. Priority Date: 09/14/2001
• Status: Active Grant

• Current Assignee: ResMed Sensor Technologies Limited (ResMed Incorporated)
• Original Assignee: University College Dublin
• Inventors: Chazal, Phillip de, sheridan, Elaine, Heneghan, Conor
• Granted Patent: US 7,025,729 B2
• Time in Patent Office: Days
• US Class Current: 600/509
• CPC Class Codes: A61B 5/335 using integrated circuit memory devices
• 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 5/7203 for noise prevention, reduction or removal
• CPC Class Codes: A61B 5/7207 of noise induced by motion artifacts
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16Z 99/00 Subject matter not provided for in other main groups of this subclass

Abstract

There is provided a method of determining a diagnostic measure of sleep apnea including the following steps: acquiring an electrocardiogram signal, calculating a set of RR intervals and electrocardiogram-derived respiratory signal from said electrocardiogram, and hence calculating a set of spectral and time-domain measurements over time periods including power spectral density, mean, and standard deviation. These measurements are processed by a classifier model which has been trained on a pre-existing data base of electrocardiogram signals to provide a probability of a specific time period containing apneic episodes or otherwise. These probabilities can be combined to form an overall diagnostic measure. The system also provides a system and apparatus for providing a diagnostic measure of sleep apnea.

Claims

1. A diagnostic recording apparatus comprising:

means for measuring cardiac electrical potential from a human for generating an electrocardiogram signal;

means for analysing said electrocardiogram signal to produce an output signal; and

means for providing a diagnostic measure of sleep apnea based on said output signal.

2. An apparatus as claimed in claim 1, including signal processing means for filtering out unwanted interference from the electrocardiogram signal and for producing a processed electrocardiogram signal for inputting to the analysing means.

3. An apparatus as claimed in claim 2, including means for recording the processed electrocardiogram signal.

4. An apparatus as claimed in claim 2, wherein the means for analysing said processed electrocardiogram signal comprises a computer algorithm performed within said apparatus.

5. An apparatus as claimed in claim 4, including a display interface which allows direct inspection of the analysis results.

6. An apparatus as claimed in claim 2, wherein the means for analysing said processed electrocardiogram signal comprises a computer algorithm performed on an external device, and the apparatus includes means for communicating with said device.

7. An apparatus as claimed in claim 1, wherein only a single channel of electrocardiogram signal is analysed.

8. An apparatus as claimed in claim 1, wherein the apparatus includes;

means for calculating the RR time interval between successive QRS complexes from the electrocardiogram signal;

means for isolating an electrocardiogram-derived respiratory signal; and

means for obtaining measurements from the RR intervals and electrocardiogram-derived respiratory signal to provide a diagnostic measure

9. A method of obtaining a diagnostic measure of sleep apnea;

the method comprising the steps of

a. acquiring single-channel or multi-channel electrocardiogram signal from a human over a period of time;

b. filtering the signal to remove electrical interference;

c. calculating a sequence of RR intervals from the electrocardiogram signal by measuring the time interval between successive QRS complexes;

d. obtaining an electrocardiogram-derived respiratory signal;

e. partitioning the electrocardiogram signal into a set of shorter time periods;

f. obtaining measurements from the sequence of RR intervals and the electrocardiogram-derived respiratory signal for each time period;

g. for each time period, calculating the probability that it can be classified as apneic or normal by processing the measurements for that time period using a classifier model which has been trained on a pre-existing database of signals; and

h combining probabilities for each time period to provide an overall diagnostic measure

10. The method of claim 9 wherein the measurements in step (f) are selected from the group consisting of:

I. the power spectral density of the electrocardiogram-derived respiratory signal;

II. the mean and standard deviation of the electrocardiogram-derived respiratory signal;

III. the power spectral density of the RR intervals;

IV. the mean and standard deviation of the RR intervals;

V. the first five serial correlation coefficients of the RR intervals;

VI. the Allan factor A(T) evaluated at a time scale T of 5, 10, 15, 20 and 25 seconds where the Allan factor is defined as

$A<br><br>\left(T\right)=\frac{E<br><br>\left\{{\left[N_{i+1}<br><br>\left(T\right)-N_i<br><br>\left(T\right)\right]}^2\right\}}{2<br><br>E<br><br>\left\{N_{i+1}<br><br>\left(T\right)\right\}},$

 

N_i(T) is the number of R wave peaks occurring in a window of length T stretching from iT to (i+1)T, and E is the expectation operator;

VII. the NN50 measure (variant

     1), defined as number of pairs of adjacent RR intervals where the first interval exceeds the second interval by more than 50 ms;

VIII. the NN50 measure (variant

     2), defined as number of pairs of adjacent RR intervals where the second interval exceeds the first interval by more than 50 ms;

IX. two pNN50 measures, defined as each NN50 measure divided by the total number of RR intervals

X. the SDSD measures, defined as the standard deviation of the differences between adjacent RR intervals, and

XI. the RMSSD measure defined as the square root of the mean of the sum of the squares of differences between adjacent RR intervals

XII. the mean or standard deviation of the RR signal over the entire recorded electrocardiogram; and

XIII. the mean or standard deviation of the electrocardiogram-derived respiratory signal over the entire recorded electrocardiogram.

11. A method as claimed in claim 9, wherein the RR interval sequence in step (c) is obtained by first identifying QRS complexes in the electrocardiogram signal and determining the times of occurrences of the R-wave peak in the QRS complexes thereby providing a series of times at which the R-wave peaks occur, followed by measuring the time interval between successive R-wave peaks to obtain the RR intervals.

12. A method as claimed in claim 9, wherein the electrocardiogram-derived respiratory signal in step (d) is obtained by extracting QRS complexes in the electrocardiogram signal and determining the times of occurrences of the R-wave peak in the QRS complexes and by taking the locations of the R-wave peaks and calculating the area represented by the signal near said locations

13. The method of claim 9 wherein after step (g), the method further includes the step of recalculating the probability of a given time period containing an apneic episode using the probabilities generated for neighbouring time periods calculated in step (g).

14. The method of claim 9 wherein the method includes the step of converting the diagnostic measure to a Apnea-Hyponea Index or Respiratory Disturbance Index.

15. A method as claimed in claim 9, wherein after step (b), the method also includes the following steps:

determining whether the electrocardiogram signal is suitable for proceeding to carrying out the analysis; and

indicating if the electrocardiogram signal is deemed unsuitable following said determination step.

16. A method as claimed in claim 10, wherein the power spectral density of the electrocardiogram-derived respiratory signal is calculated using an averaged periodogram technique.

17. A method as claimed in claim 10, wherein the power spectral density of the RR intervals is calculated using an averaged periodogram technique

18. A method as claimed in claim 9 in which the classifier model is selected from the group consisting of:

(1) a linear discriminant classifier;

(2) a quadratic discriminant classifier; and

(3) a regularised quadratic discriminant classifier.

19. A system for providing a diagnostic measure for sleep apnea including:

a diagnostic recording apparatus including means for measuring cardiac potentials from a human to generate unprocessed electrocardiogram signal;

means for processing the unprocessed electrocardiogram signal and extracting RR intervals and an electrocardiogram-derived respiratory signal from the unprocessed electrocardiogram signal, where RR intervals are defined as the time durations between successive QRS complexes in the electrocardiogram signal;

means for dividing the electrocardiogram signal into shorter time periods;

means for identifying time periods as containing apneic episodes and time periods not containing apneic episodes;

means for providing reports which indicate apneic periods, non-apneic periods and indeterminate periods, and for correlating these events with the electrocardiogram signal; and

means for providing an overall diagnostic measure in a clinically meaningful form based on the reports.

20. A system as claimed in claim 19, including means for correlating the apneic, non-apneic and indeterminate periods with the electrocardiogram signal using a visual representation.