Heart sound tracking system and method

• US 9,364,193 B2
• Filed: 06/03/2015
• Issued: 06/14/2016
• Est. Priority Date: 04/17/2007
• Status: Active Grant

• Current Assignee: Cardiac Pacemakers Incorporated (Boston Scientific Corporation)
• Original Assignee: Cardiac Pacemakers Incorporated (Boston Scientific Corporation)
• Inventors: Patangay, Abhilash, Siejko, Krzysztof Z., Carlson, Gerrard M.
• Primary Examiner(s): Layno, Carl H
• Assistant Examiner(s): Morales, Jon Eric C
• Application Number:
  US14/729,187
• Publication Number:
  US 20150257729A1
• Time in Patent Office: 377 Days
• Field of Search: 607/17
• US Class Current: 1/1
• CPC Class Codes: A61B 7/00 Instruments for auscultation
• CPC Class Codes: A61B 7/04 Electric stethoscopes microphones, acoustic transducers therefor H04R
• CPC Class Codes: A61N 1/3627 for treating a mechanical deficiency of the heart, e.g. congestive heart failure or cardiomyopathy
• CPC Class Codes: A61N 1/36514 controlled by a physiological quantity other than heart potential, e.g. blood pressure controlled by two or more physical parameters A61N1/36585
• CPC Class Codes: A61N 1/36521 the parameter being derived from measurement of an electrical impedance
• CPC Class Codes: A61N 1/36585 controlled by two or more physical parameters
• CPC Class Codes: A61N 1/368 comprising more than one electrode co-operating with different heart regions A61N1/3622, A61N1/3627 take precedence

Abstract

A system and method provide heart sound tracking, including an input circuit, configured to receive heart sound information, and a heart sound recognition circuit. The heart sound recognition circuit can be coupled to the input circuit and can be configured to recognize, within a particular heart sound of a particular heart sound waveform, a first intra heart sound energy indication and a corresponding first intra heart sound time indication using the heart sound information from the particular heart sound waveform and the heart sound information from at least one other heart sound waveform. The particular heart sound can include at least a portion of one of S1, S2, S3, and S4. Further, the first intra heart sound energy indication and the corresponding first intra heart sound time indication can correspond to the at least a portion of one of S1, S2, S3, and S4, respectively.

Claims

1. A method for identifying a heart sound in a heart sound waveform, the method comprising:

identifying one or more first candidate heart sounds in a first heart sound waveform, each of the one or more first candidate heart sounds including corresponding first time and first energy features;

identifying two or more second candidate heart sounds in a second heart sound waveform, each of the two or more second candidate heart sounds including corresponding second time and second energy features; and

using a cost analysis circuit;

determining a temporal cost for each of the second time features of the two or more second candidate heart sounds relative to each of the first time features of the one or more first candidate heart sounds;

determining a score for each of the two or more second candidate heart sounds using the determined temporal cost for each candidate; and

discriminating between the two or more second candidate heart sounds, based on the determined scores, to provide an identifier indicating a detected heart sound.

2. The method of claim 1, further comprising using the cost analysis circuit to determine a local benefit for each of the second energy features of the two or more second candidate heart sounds relative to each of the first energy features of the one or more first candidate heart sounds;

wherein the determining the score for each of the two or more second candidate heart sounds includes using the determined temporal cost and the determined local benefit for each candidate.

3. The method of claim 2, further comprising identifying a heart sound window in a heart sound signal from a heart sound sensor, the heart sound window corresponding to multiple occurrences of S1, S2, S3, or S4 heart sounds, wherein the identifying the one or more first candidate heart sounds and the identifying the two or more second candidate heart sounds includes using a portion of the second heart sound waveform that corresponds to the heart sound window.

4. The method of claim 2, wherein the identifying the two or more second candidate heart sounds in a second heart sound waveform includes using a heart sound waveform that is sequentially adjacent to the first heart sound waveform.

5. The method of claim 2, wherein the identifying the one or more first candidate heart sounds in the first heart sound waveform includes using an ensemble average of multiple heart sound waveforms.

6. The method of claim 5, further comprising generating an ensemble average of multiple heart sound waveforms corresponding to multiple consecutive cardiac cycles.

7. The method of claim 2, wherein the determining the cost for each of the two or more second candidate heart sounds includes identifying a maximum of a difference between a previous score and a temporal cost for each of the second time features of the two or more second candidate heart sounds.

8. The method of claim 7, wherein the determining the cost for each of the two or more second candidate heart sounds includes adding the local benefit to the identified maximum.

9. The method of claim 8, wherein the determining the cost includes using a scaling factor for at least one of the temporal cost and the local benefit.

10. The method of claim 2, further comprising determining a back pointer that indicates a preferred one of the one or more first candidate heart sounds based on the identifier indicating the detected heart sound.

11. The method of claim 10, further comprising identifying two or more third candidate heart sounds in a third heart sound waveform, each of the two or more third candidate heart sounds including corresponding third time and third energy features;

determining a temporal cost for each of the third time features of the two or more third candidate heart sounds relative to each of the second time features of the two or more second candidate heart sounds;

wherein the determining the back pointer includes identifying a maximum of a difference between the score for each of the two or more second candidate heart sounds and the temporal cost for each of the third time features of the two or more third candidate heart sounds.

12. The method of claim 1, wherein the identifying the one or more first candidate heart sounds in the first heart sound waveform includes identifying the corresponding first time and first energy features corresponding to peaks of the first heart sound waveform.

13. The method of claim 1, wherein the discriminating between the two or more second candidate heart sounds to provide the identifier includes providing an identifier of an S3 heart sound.

14. The method of claim 13, further comprising using the identifier of the S3 heart sound, together with subsequent heart sound waveform information, to provide an indication of heart failure.

15. A method for iteratively providing a heart sound identifier for discriminating a heart sound of a specified heart sound type in a heart sound waveform that is generated by a heart sound sensor, the method comprising

identifying one or more first candidate heart sounds in a first heart sound waveform, each of the one or more first candidate heart sounds including corresponding first time and first peak energy features;

identifying one or more second candidate heart sounds in a second heart sound waveform, each of the one or more second candidate heart sounds including corresponding second time and second peak energy features; and

identifying one or more third candidate heart sounds in a third heart sound waveform, each of the one or more third candidate heart sounds including corresponding third time and third peak energy features; and

using a cost analysis circuit;

determining a temporal cost and a local benefit for each of the second time features and the second peak energy features of the one or more second candidate heart sounds relative to each of the first time features and the first peak energy features of the one or more first candidate heart sounds, respectively;

determining a temporal cost and a local benefit for each of the third time features and the third peak energy features of the one or more third candidate heart sounds relative to each of the second time features and the second peak energy features of the one or more second candidate heart sounds, respectively;

determining a score for each of the one or more second candidate heart sounds using the determined temporal cost and local benefit for each candidate;

determining a score for each of the one or more third candidate heart sounds using the determined temporal cost and local benefit for each candidate; and

discriminating between the second and third candidate heart sounds, based on the determined scores, to provide an identifier indicating a detected heart sound in each of the second and third heart sound waveforms.

16. The method of claim 15, wherein the identifying the first, second, or third candidate heart sounds includes using an ensemble average of multiple heart sounds of the same heart sound type.

17. A system, comprising:

an input circuit configured to receive heart sound waveform information from a heart sound sensor; and

a heart sound recognition circuit, coupled to the input circuit, configured to recognize;

one or more first candidate heart sounds in a first portion of the heart sound waveform information, each of the one or more first candidate heart sounds including first time and corresponding first peak energy features; and

two or more second candidate heart sounds in a second portion of the heart sound waveform information, each of the two or more second candidate heart sounds including second time and corresponding second peak energy features; and

a cost analysis circuit configured to determine a temporal cost for each of the second time features of the two or more second candidate heart sounds relative to each of the first time features of the one or more first candidate heart sounds;

wherein the cost analysis circuit is configured to discriminate between the two or more second candidate heart sounds, based on the determined temporal costs, to provide an identifier indicating a detected heart sound.

18. The system of claim 17, wherein the cost analysis circuit is further configured to determine a local benefit for each of the second peak energy features of the two or more second candidate heart sounds relative to each of the first peak energy features of the one or more first candidate heart sounds;

wherein the cost analysis circuit is further configured to determine a score for each of the two or more second candidate heart sounds using the determined temporal cost and the determined local benefit; and

wherein the cost analysis circuit is further configured to discriminate between the two or more second candidate heart sounds, based on the determined score, to provide the identifier indicating the detected heart sound.

19. The system of claim 18, further comprising a coasting circuit configured to disregard the identifier indicating the detected heart sound when the determined score exceeds a first specified threshold score or when the determined score is less than a second specified threshold score.

20. The system of claim 19, further comprising a recapture circuit configured to recapture the disregarded identifier based on a comparison of the identifier with a specified time feature and a corresponding specified peak energy feature of a known heart sound.