Ventricular activation (RR) entropy change as a predictor of sudden cardiac death in cardiac resynchronization therapy patients

• US 9,839,364 B2
• Filed: 05/11/2013
• Issued: 12/12/2017
• Est. Priority Date: 10/06/2009
• Status: Active Grant

• Current Assignee: Johns Hopkins University, University of Virginia Patent Foundation (University of Virginia)
• Original Assignee: Johns Hopkins University, University of Virginia Patent Foundation (University of Virginia)
• Inventors: Moorman, J. Randall, Lake, Douglas E., Tomaselli, Gordon F., Demazumder, Deeptankar
• Primary Examiner(s): DIETRICH, JOSEPH M
• Application Number:
  US14/400,408
• Publication Number:
  US 20150133807A1
• Time in Patent Office: 1,676 Days
• CPC Class Codes: A61B 5/0006 ECG or EEG signals
• CPC Class Codes: A61B 5/02405 Determining heart rate variability
• CPC Class Codes: A61B 5/316 Modalities, i.e. specific diagnostic methods
• CPC Class Codes: A61B 5/33 specially adapted for cooperation with other devices
• CPC Class Codes: A61B 5/346 Analysis of electrocardiograms
• CPC Class Codes: A61B 5/352 Detecting R peaks, e.g. for synchronising diagnostic apparatus; Estimating R-R interval

Abstract

A method of determining health and mortality includes obtaining a ventricular activation (RR) time series from a subject for multiple temporal intervals. The method also includes calculating a cardiac entropy in the RR time series over the temporal intervals using coefficient of sample entropy (COSEn). Additionally, the method includes comparing the cardiac entropy between the intervals to determine health and mortality. The absolute and relative changes in entropy over a patient'"'"'s follow up period provide dynamic information regarding health and mortality risk. The determination of health and mortality can then be used to create a treatment plan for the subject.

Claims

1. A method for determining increased risk of mortality of a patient, comprising:

measuring a first cardiac rhythm of a patient at a first time;

identifying at least one segment of said first cardiac rhythm;

calculating a first entropy measurement for the at least one segment;

measuring a second cardiac rhythm of said patient at a second time subsequent to said first time;

identifying at least one segment of said second cardiac rhythm;

calculating a second entropy measurement for the at least one segment of said second cardiac rhythm;

calculating a change in entropy by comparing said second entropy measurement with said first entropy measurement; and

determining that said patient is at an increased risk of mortality when said change in entropy is above a predetermined value.

2. The method of claim 1, wherein the entropy measurement is a measurement of absolute entropy.

3. The method of claim 2, wherein the absolute entropy measurement is a coefficient of sample entropy (COSEn).

4. The method of claim 3, wherein the at least one cardiac rhythm arises from at least a deterministic process.

5. The method of claim 3, wherein the at least one cardiac rhythm arises from a combination of both deterministic and stochastic physiological processes.

6. The method of claim 3, wherein the at least one segment comprises a series of beats having a statistically homogeneous time interval between beats.

7. The method of claim 4, wherein each at least one cardiac rhythm comprises a heart rate time series such as would be provided by non-invasive devices that do not use a conventional ECG signal.

8. The method of claim 3, wherein the heart rate time series comprises a number of beats, and wherein COSEn is calculated at least every 50 beats.

9. The method of claim 3, wherein the at least one cardiac rhythm comprises an RR-interval series, and wherein the step of calculating COSEn for at least one segment comprises:

calculating a mean RR-interval for the RR-interval series;

using the mean RR-interval as a continuous variable;

unit mean normalizing the RR-interval series by dividing each observation by the mean RR-interval; and

calculating COSEn as an entropy rate or entropy of the unit mean normalized RR interval series.

10. The method of claim 3, wherein the at least one cardiac rhythm comprises an RR-interval series, and wherein the step of calculating COSEn for at least one segment comprises:

calculating the differential quadratic entropy rate using a sample entropy (SampEn) algorithm;

calculating a mean RR-interval for the RR-interval series; and

subtracting the natural logarithm of the mean RR-interval from the differential quadratic entropy rate to obtain COSEn.

11. The method of claim 1, wherein said first cardiac rhythm is measured when said patient is in normal sinus rhythm (NSR).

12. The method of claim 1, wherein said second time is on the order of months subsequent to said first time.

13. The method of claim 1, wherein calculation of a change in entropy comprises calculating a rate of entropy change.

14. The method of claim 13, wherein calculation of a rate of entropy change comprises measuring a slope of a linear regression fit to values at a baseline measurement and subsequent measurements of cardiac rhythm.

15. The method of claim 11, wherein said change in entropy is a rising entropy of NSR.

16. The method of claim 1, wherein determining that said patient is at increased risk of mortality further comprises use of a multivariable model that employs entropy measures.

17. An apparatus comprising:

a programmable computer, programmed to

measure a first cardiac rhythm of a patient at a first time;

identify at least one segment of said first cardiac rhythm;

calculate a first entropy measurement for the at least one segment;

measure a second cardiac rhythm of said patient at a second time subsequent to said first time;

identify at least one segment of said second cardiac rhythm;

calculate a second entropy measurement for the at least one segment of said second cardiac rhythm; and

calculate a change in entropy by comparing said second entropy measurement with said first entropy measurement.

18. The apparatus of claim 17, wherein the entropy measurement is a measurement of absolute entropy.

19. The apparatus of claim 18, wherein the absolute entropy measurement is a coefficient of sample entropy (COSEn).

20. The apparatus of claim 19, wherein the at least one cardiac rhythm comprises an RR-interval series, and wherein calculating COSEn for at least one segment comprises:

calculating a mean RR-interval for the RR-interval series;

using the mean RR-interval as a continuous variable;

unit mean normalizing the RR-interval series by dividing each observation by the mean RR-interval; and

calculating COSEn as an entropy rate or entropy of the unit mean normalized RR interval series.

21. The apparatus of claim 19, wherein the at least one cardiac rhythm comprises an RR-interval series, and wherein calculating COSEn for at least one segment comprises:

calculating the differential quadratic entropy rate using a sample entropy (SampEn) algorithm;

calculating a mean RR-interval for the RR-interval series; and

subtracting the natural logarithm of the mean RR-interval from the differential quadratic entropy rate to obtain COSEn.

22. The apparatus of claim 17, wherein said first cardiac rhythm is measured when said patient is in normal sinus rhythm (NSR).

23. The apparatus of claim 17, wherein said second time is on the order of months subsequent to said first time.

24. The apparatus of claim 17, wherein calculation of a change in entropy comprises calculating a rate of entropy change.

25. The apparatus of claim 24, wherein calculation of a rate of entropy change comprises measuring a slope of a linear regression fit to values at a baseline measurement and subsequent measurements of cardiac rhythm.

26. The apparatus of claim 22, wherein said change in entropy is a rising entropy of NSR.

27. A computer program product comprising a non-transient computer readable storage medium storing computer-executable instructions causing a computer to:

receive a measurement of a first cardiac rhythm of a patient taken at a first time;

identify at least one segment said first cardiac rhythm;

calculate a first entropy measurement for the at least one segment;

receive a measurement of a second cardiac rhythm of said patient taken at a second time subsequent to said first time;

identify at least one segment of said second cardiac rhythm;

calculate a second entropy measurement for the at least one segment of said second cardiac rhythm; and

calculate a change in entropy by comparing said second entropy measurement with said first entropy measurement.