System and method for classifying cardiac complexes

• US 20020193696A1
• Filed: 08/14/2002
• Published: 12/19/2002
• Est. Priority Date: 02/12/1999
• Status: Active Grant

• Current Assignee: Cardiac Pacemakers Incorporated (Boston Scientific Corporation)
• Original Assignee: Cardiac Pacemakers Incorporated (Boston Scientific Corporation)
• Inventors: Hsu, William, Smith, Joseph Martin
• Granted Patent: US 6,728,572 B2
• Time in Patent Office: Days
• US Class Current: 600/512
• CPC Class Codes: A61B 5/35 by template matching
• 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: A61N 1/3622 comprising two or more electrodes co-operating with different heart regions
• CPC Class Codes: A61N 1/3704 Circuits specially adapted therefor, e.g. for sensitivity control
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

A system and method classify cardiac complexes based on cardiac information derived from two or more cardiac signals. Two or more cardiac signals containing cardiac complexes are monitored. A cardiac complex in the two or more cardiac signals is isolated in an analysis window. The morphology of the cardiac complex in each of the two or more cardiac signals is then compared to the morphology of a template cardiac complex representing a predetermined cardiac condition. Based on this comparison, the cardiac complex is classified as either belonging or not belonging to the predetermined cardiac condition.

Claims

1. A method, comprising:

detecting a first cardiac complex in a first cardiac signal and a second cardiac complex in a second cardiac signal, the first and second cardiac complexes associated with the same heart contraction;

comparing the first cardiac complex to a first template cardiac complex and the second cardiac complex to a second template cardiac complex to determine whether the heart contraction is representative of a predetermined cardiac rhythm state represented by the first and second template cardiac complexes; and

classifying whether heart contraction is representative of the predetermined cardiac rhythm state based on the comparison of the first cardiac complex to the first template cardiac complex and the second cardiac complex to the second template cardiac complex.

2. The method of claim 1, further comprising determining and delivering a therapy at least in part using the classification of the heart contraction.

3. The method of claim 1, wherein the predetermined cardiac rhythm state includes at least one of:

a normal sinus rhythm (NSR);

a ventricular tachycardia (VT);

a supraventricular tachycardia (SVT);

a premature ventricular contraction (PVC); and

an ischemic event.

4. The method of claim 1, wherein the first and second template cardiac complexes are template morphologies and comparing the first and second cardiac complexes to the corresponding first and second template cardiac complexes comprises comparing morphologies of the first and second cardiac complexes to the respective first and second template morphologies.

5. The method of claim 4, wherein comparing the morphologies comprises comparing at least one predetermined feature in the first and second cardiac complexes to a corresponding at least one predetermined feature in the corresponding one of the first and second template morphologies.

6. The method of claim 5, further comprising aligning the predetermined feature in the first and second cardiac complexes and the respective first and second morphologies.

7. The method of claim 4, in which the comparing comprises computing a correlation between the first and second cardiac complexes and the respective first and second morphologies.

8. A method, comprising:

detecting a first cardiac complex in a first cardiac signal and a second cardiac complex in a second cardiac signal, the first and second cardiac complexes associated with the same heart contraction;

locating at least one first predetermined feature in the first cardiac complex and at least one second predetermined feature in the second cardiac complex;

generating at least one characteristic quantity using at least one of the first and second predetermined features; and

classifying whether a heart contraction represents a predetermined cardiac rhythm state based at least in part on the characteristic quantity.

9. The method of claim 8, further comprising determining and delivering a therapy at least in part in response to the classification of the heart contraction.

10. The method of claim 8, further including monitoring the first and second cardiac signals to provide a substantially simultaneous view of a cardiac complex from at least two different cardiac locations.

11. The method of claim 8, further comprising:

creating at least one classification quantity, wherein the classification quantity represents a predetermined cardiac rhythm state; and

comparing the characteristic quantity to the classification quantity to determine whether the characteristic quantity is representative of the predetermined cardiac rhythm state;

wherein the classifying the heart contraction is based at least in part on the outcome of the comparing the characteristic quantity to the classification quantity.

12. The method of claim 11, wherein the characteristic quantity comprises a characteristic scalar value, the classification quantity comprises a classification scalar value, and the comparing the characteristic quantity to the classification quantity comprises comparing the characteristic scalar value to the classification scalar value.

13. The system of claim 12, wherein the characteristic scalar value is representative, at least in part, of at least one of:

a time associated with at least one of the first and second predetermined features;

a time difference between the first and second predetermined features;

an amplitude value of at least one of the first and second predetermined features; and

a time difference between two predetermined features in one cardiac signal.

14. The method of claim 11, wherein:

the characteristic quantity is a characteristic vector comprising two or more scalar values;

the classification quantity is a classification vector; and

comparing the characteristic quantity to the classification quantity comprises comparing the characteristic vector to the classification vector.

16. The method of claim 15, in which the predetermined feature comprises a maximum deflection of the first signal along an R-wave in the cardiac complex detected in the first signal and present in the first NSR representative complex.

17. The method of claim 15, in which the comparing comprises determining a degree of correlation between morphological features in the second signal and the second NSR representative complex.

18. The method of claim 15, in which the comparing comprises comparing an amplitude distribution of morphological features in the second signal to an amplitude distribution of morphological features in the second NSR representative complex.

19. The method of claim 15, in which the comparing comprises comparing a spectral analysis of the cardiac complex detected in the second signal to a spectral analysis of the second NSR representative complex.

20. The method of claim 15, in which the monitoring the first signal includes sensing electrical cardiac activity through a near-field sensing channel, and sensing the second signal includes sensing electrical cardiac activity through a far-field sensing channel.

21. The method of claim 15, in which the comparing includes comparing to a second NSR representative complex indicative of a ventricular tachycardia (VT) complex.

22. The method of claim 21, in which the classifying includes classifying, over a plurality of detected cardiac complexes, a percentage of the detected cardiac complexes as VT complexes, and further comprising applying therapy to treat VT if the percentage exceeds a predetermined threshold.

24. The method of claim 23, in which the locating the at least one predetermined feature includes locating at least one of:

a maximum deflection;

a predetermined deviation from a baseline signal, the predetermined deviation indicating a beginning of the cardiac complex; and

, a return to a baseline, for a predetermined time period, the return to baseline indicating an ending of the cardiac complex.

25. The method of claim 23, in which the scalar is a time of occurrence of the at least one predetermined feature.

26. The method of claim 23, in which the scalar is a difference in time between the occurrence of two of the at least one predetermined features.

27. The method of claim 23, in which the scalar is a signal amplitude of the at least one predetermined feature.

28. The method of claim 23, in which the comparing includes performing a correlation between the cardiac vector and the one or more classification vectors.

29. The method of claim 23, further comprising aligning the cardiac vector with each of the one or more classification vectors before the comparing the cardiac vector.

30. The method of claim 29, in which the aligning comprises adjusting each scalar value of the cardiac vector such that a value of one of the scalars in the cardiac vector equals a value of a corresponding scalar in the classification vector.

15. A method, comprising:

monitoring a first signal and a second signal representative of electrical cardiac activity;

detecting a cardiac complex in each of the first signal and the second signal;

locating a predetermined feature in the cardiac complex detected in the first signal and present in a first NSR representative complex;

aligning the predetermined feature in the cardiac complex detected in the first signal and the first NSR representative complex;

comparing the cardiac complex detected in the second signal to a second NSR representative complex; and

classifying whether the cardiac complex is an arrhythmic cardiac complex based on the comparing of the cardiac complex in the second cardiac signal to the second NSR representative complex.

23. A method, comprising:

monitoring two or more cardiac signals representative of electrical cardiac activity;

detecting a cardiac complex in each of the two or more cardiac signals;

locating at least one predetermined feature in the cardiac complex for each of the two or more cardiac signals;

generating a scalar value representative of the at least one predetermined feature for each of the two or more cardiac signals;

creating a cardiac vector using the scalar values, where the cardiac vector represents the cardiac complex;

comparing the cardiac vector to one or more classification vectors; and

classifying the cardiac complex based on the comparing the cardiac vector to one or more classification vectors.

31. A system, comprising:

a sensing system, including first and second channels to respectively sense a cardiac complex present in first and second signals representative of electrical cardiac activity;

a morphology analyzer circuit, coupled to the sensing system, wherein the morphology analyzer circuit is configured to locate at least one predetermined feature in a morphology of the cardiac complex in the first signal and the second signal, and to use the at least one predetermined feature to generate, for each of the first and second signals, a cardiac complex vector, the cardiac complex vector including two or more scalar values; and

a signal feature comparison circuit, coupled to the morphology analyzer circuit, to compare the cardiac complex vector to one or more classification vectors to classify the cardiac complex.

32. The system of claim 31, in which the morphology analyzer circuit is configured to located at least one predetermined feature that includes at least one of:

a maximum deflection of the first signal and the second signal;

a predetermined deviation of the of the first signal and the second signal from a baseline signal indicating a beginning of the cardiac complex; and

a return of the first signal and the second signal to a baseline signal for a predetermined time period indicating an ending of the cardiac complex.

33. The system of claim 31, in which at least one of the two or more scalar values represents a time the at least one predetermined feature in the morphology of the cardiac complex occurred.

34. The system of claim 31, in which at least one of the two or more scalar values represents a difference in time between occurrences of a first predetermined feature and a second predetermined feature.

35. The system of claim 31, in which at least one of the two or more scalar values represents a signal amplitude of the at least one predetermined feature of the first signal and the second signal.

36. The system of claim 31, in which the signal feature comparison circuit is configured to align the cardiac complex vector with each of the one or more classification vectors, wherein the cardiac complex vector is aligned by adjusting each scalar value of the cardiac complex vector such that one of the scalars in the cardiac complex vector is substantially equal in value to a corresponding scalar in a classification vector.

38. The system of claim 37, in which the signal feature comparison circuit is configured to align the predetermined feature comprising a maximum deflection along an R-wave, and configured to then compare the cardiac complex of the second signal.

39. The system of claim 37, in which the sensing system includes a near-field sensing channel over which the first signal is monitored and a far-field sensing channel over which the second signal is monitored.

40. The system of claim 37, in which the signal feature comparison circuit is configured to determine a percentage of ventricular tachycardia (VT) complexes over a plurality of cardiac complexes, and further comprising a control circuit configured to apply VT therapy when the percentage of ventricular tachycardia complexes exceeds a predetermined threshold.

37. A system, comprising:

a sensing system, configured to sense first and second signals representative of electrical cardiac activity;

a template generator circuit, coupled to the sensing system, configured to determine first and second NSR representative complexes from the first and second signals, respectively, sensed during normal sinus rhythm;

a morphology analyzer circuit, coupled to the sensing system, configured to locate a predetermined feature in a cardiac complex detected in the first signal and in the first NSR representative complex; and

a signal feature comparison circuit, coupled to the morphology analyzer circuit and the template generator circuit, configured to align the predetermined feature in the cardiac complex of the first signal with the predetermined feature of the first NSR representative complex, and configured to then compare the cardiac complex of the second signal to the second NSR representative complex.