Publication NumberUS 7912535
Assignees
  • University of Rochester
StatusIssued Patent
Application Number11/680896
AvailabilityUnknown
Filing Date2007-03-01
Publication Date2011-03-22

Abstract

A method for assessing repolarization abnormalities is disclosed. At least two repolarization signals are identified from a set of ECG signals. PCA analysis is performed on the at least two repolarization signals to extract at least eigenvectors ev1 and ev2. A maximum vector MV is determined based on a transformed ECG signal in a plane defined by ev1 and ev2. A repolarization duration is determined which is based on the maximum vector MV. A system for assessing repolarization abnormalities is also disclosed. The system has a processor configured to determine a repolarization duration which is based on a maximum vector MV from transformed ECG repolarization signals in a plane defined by eigenvectors ev1 and ev2 which result from PCA analysis on the ECG repolarization signals. The system also has a data input coupled to the processor and a user interface coupled to either the processor or the data input.

Claims

  • 1. A method for assessing repolarization abnormalities, comprising: using a processor, identifying at least two repolarization signals from a set of electrocardiogram (ECG) signals; using the processor, performing principal component analysis on the at least two repolarization signals to extract at least eigenvectors ev1 and ev2; using the processor, determining a maximum vector MV based on a transformed ECG signal in a plane defined by ev1 and ev2; and using the processor, determining a repolarization duration which is based on the maximum vector MV.
  • 2. The method of claim 1, further comprising: filtering the set of ECG signals.
  • 3. The method of claim 2, wherein filtering the set of ECG signals comprises low-pass FIR filtering the ECG signals.
  • 4. The method of claim 2, wherein filtering the set of ECG signals comprises removing a wandering baseline.
  • 5. The method of claim 2, further comprising: detecting at least one QRS complex in the set of ECG signals.
  • 6. The method of claim 5, wherein identifying at least two repolarization signals from the set of ECG signals comprises defining each of the at least two repolarization signals as starting from a point J and ending at a point relative to a following R peak.
  • 7. The method of claim 6, wherein the point relative to the following R peak comprises a point which is an arbitrary point in time prior to the following R peak.
  • 8. The method of claim 7, wherein the arbitrary point in time prior to the following R peak comprises a point which is approximately 220 milliseconds prior to the following R peak.
  • 9. The method of claim 2, wherein filtering the set of ECG signals comprises statistically combining multiple beats from the ECG signals.
  • 10. The method of claim 9, wherein statistically combining multiple beats from the ECG signals comprises creating a set of median beats, each median beat from one of the ECG signals.
  • 11. The method of claim 2, wherein filtering the set of ECG signals comprises discarding one or more leading beats from the ECG signals.
  • 12. The method of claim 2, wherein filtering the set of ECG signals comprises discarding one or more trailing beats from the ECG signals.
  • 13. The method of claim 2, wherein filtering the set of ECG signals comprises discarding beats which do not have a corresponding stable heart rate.
  • 14. The method of claim 13, wherein discarding beats which do not have a corresponding stable heart rate comprises discarding beats which have a heart rate that varies by more than a certain percentage in a previous arbitrary time frame.
  • 15. The method of claim 14, wherein the certain percentage is ten percent and the arbitrary time frame is two minutes.
  • 16. The method of claim 1, wherein principal component analysis on the at least two repolarization signals to extract at least eigenvectors ev1 and ev2 comprises: subtracting a mean from the at least two repolarization signals to produce a data set with a mean of zero; calculating a covariance matrix for the data set; calculating a set of eigenvectors and corresponding eigenvalues for the covariance matrix; determining that a first eigenvector from the set of eigenvectors with the largest corresponding eigenvalue is the eigenvector ev1; and determining that a second eigenvector from the set of eigenvectors with the second-largest corresponding eigenvalue is the eigenvector ev2.
  • 17. The method of claim 1, wherein determining the maximum vector MV comprises determining MV=max{right arrow over (VECG)}(t)−{right arrow over (VECG)}(TQ), where {right arrow over (VECG)}(t)={right arrow over (ev)}1(t)+{right arrow over (ev)}2 (t), and TQ is a time coinciding with a beginning of the QRS complex.
  • 18. The method of claim 17, wherein determining the repolarization duration which is based on the maximum vector MV comprises determining an early repolarization duration (ERD) at a threshold percentage x % of MV, such that: ERDx %=TMV−TE, where TE is a value for t where the following equation is fulfilled: ∥VECG(t)−VECG(TMV)=MV·x %∥, with t
  • 19. The method of claim 17, wherein determining the repolarization duration which is based on the maximum vector MV comprises determining a late repolarization duration (LRD) at a threshold percentage x % of MV, such that: LRDx %=TL−TMV, where TL is a value for t where the following equation is fulfilled: ∥VECG(t)−VECG(TMV)=MV·x %∥, with t>TMV.
  • 20. The method of claim 17, wherein determining the repolarization duration which is based on the maximum vector MV comprises determining: a) an early repolarization duration (ERD) at a first threshold percentage x1% of MV, such that: ERDx1%=TMV−TE, where TE is a value for t where the following equation is fulfilled: ∥VECG(t)−VECG(TMV)=MV·x1%∥, with tTMV; and c) a total repolarization duration (TRD) such that TRD=ERDx1%+LRDx2%.
  • 21. The method of claim 20, wherein the first threshold percentage x1% and the second threshold percentage x2% are equal.
  • 22. The method of claim 20, wherein the first threshold percentage x1% and the second threshold percentage x2% are not equal.
  • 23. The method of claim 1, further comprising: constructing a T-loop based on a plot of a transformed ECG signal in a plane defined by ev1 and ev2.
  • 24. The method of claim 23, wherein determining the maximum vector MV comprises determining a vector having a largest magnitude from a first sample point in the T-loop to a second sample point in the T-loop.
  • 25. The method of claim 1, wherein determining the repolarization duration which is based on the maximum vector MV comprises: taking a threshold percentage of a magnitude of MV to determine a threshold magnitude; and projecting the threshold magnitude relative to an endpoint of MV to determine a beginning point and an ending point on the transformed ECG signal in the plane defined by ev1 and ev2.
  • 26. The method of claim 25, wherein: projecting the threshold magnitude around the endpoint of MV to determine the beginning point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the beginning point as a first sample point of the transformed ECG signal, the beginning point being a point which is closest to the projected threshold magnitude on a first side of the endpoint of MV; and projecting the threshold magnitude around the endpoint of MV to determine the ending point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the ending point as a second sample point of the transformed ECG signal, the ending point being a point which is closest to the projected threshold magnitude on a second side of the endpoint of MV.
  • 27. The method of claim 25, wherein: projecting the threshold magnitude around the endpoint of MV to determine the beginning point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the beginning point as an interpolated beginning point between a first sample point and a second sample point of the transformed ECG signal on a first side of the endpoint of MV, the first and second sample points being the two closest points to the projected threshold magnitude on the first side of the endpoint of MV; and projecting the threshold magnitude around the endpoint of MV to determine the ending point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the ending point as an interpolated ending point between a third sample point and a fourth sample point of the transformed ECG signal on a second side of the endpoint of MV, the third and fourth sample points being the two closest points to the projected threshold magnitude on the second side of the endpoint of MV.
  • 28. The method of claim 25, wherein determining the repolarization duration which is based on the maximum vector MV comprises: determining a beginning time corresponding to the beginning point; determining a peak time corresponding to the endpoint of MV; and determining an early repolarization duration (ERD) for the threshold percentage which comprises a time difference between the peak time and the beginning time.
  • 29. The method of claim 28, wherein determining the repolarization duration which is based on the maximum vector MV further comprises: determining an ending time corresponding to the ending point; and determining a late repolarization duration (LRD) for the threshold percentage which comprises a time difference between the ending time and the peak time.
  • 30. The method of claim 29, wherein determining the repolarization duration which is based on the maximum vector MV further comprises: determining a total repolarization duration (TRD) for the threshold percentage comprising the sum of ERD and LRD.
  • 31. The method of claim 1, wherein determining the repolarization duration which is based on the maximum vector MV comprises: taking a threshold percentage of a magnitude of MV to determine a threshold magnitude; and projecting the threshold magnitude relative to an endpoint of MV to determine a beginning point on the transformed ECG signal in the plane defined by ev1 and ev2.
  • 32. The method of claim 31, wherein: projecting the threshold magnitude around the endpoint of MV to determine the beginning point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the beginning point as a first sample point of the transformed ECG signal, the beginning point being a point which is closest to the projected threshold magnitude on a first side of the endpoint of MV.
  • 33. The method of claim 31, wherein: projecting the threshold magnitude around the endpoint of MV to determine the beginning point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the beginning point as an interpolated beginning point between a first sample point and a second sample point of the transformed ECG signal on a first side of the endpoint of MV, the first and second sample points being the two closest points to the projected threshold magnitude on the first side of the endpoint of MV.
  • 34. The method of claim 31, wherein determining the repolarization duration which is based on the maximum vector MV comprises: determining a beginning time corresponding to the beginning point; determining a peak time corresponding to the endpoint of MV; and determining an early repolarization duration (ERD) for the threshold percentage which comprises a time difference between the peak time and the beginning time.
  • 35. The method of claim 1, wherein determining the repolarization duration which is based on the maximum vector MV comprises: taking a threshold percentage of a magnitude of MV to determine a threshold magnitude; and projecting the threshold magnitude relative to an endpoint of MV to determine an ending point on the transformed ECG signal in the plane defined by ev1 and ev2.
  • 36. The method of claim 35, wherein: projecting the threshold magnitude around the endpoint of MV to determine the ending point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the ending point as a sample point of the transformed ECG signal, the ending point being a point which is closest to the projected threshold magnitude on a second side of the endpoint of MV.
  • 37. The method of claim 35, wherein: projecting the threshold magnitude around the endpoint of MV to determine the ending point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the ending point as an interpolated ending point between a first sample point and a second sample point of the transformed ECG signal on a second side of the endpoint of MV, the first and second sample points being the two closest points to the projected threshold magnitude on the second side of the endpoint of MV.
  • 38. The method of claim 37, wherein determining the repolarization duration which is based on the maximum vector MV further comprises: determining a peak time corresponding to the endpoint of MV; determining an ending time corresponding to the ending point; and determining a late repolarization duration (LRD) for the threshold percentage which comprises a time difference between the ending time and the peak time.
  • 39. The method of claim 1, wherein determining the repolarization duration which is based on the maximum vector MV comprises: taking a first threshold percentage of a magnitude of MV to determine a first threshold magnitude; taking a second threshold percentage of the magnitude of MV to determine a second threshold magnitude; projecting the first threshold magnitude relative to an endpoint of MV to determine a beginning point on the transformed ECG signal in the plane defined by ev1 and ev2; and projecting the second threshold magnitude relative to an endpoint of MV to determine an ending point on the transformed ECG signal in the plane defined by ev1 and ev2.
  • 40. The method of claim 39, wherein: projecting the first threshold magnitude around the endpoint of MV to determine the beginning point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the beginning point as a first sample point of the transformed ECG signal, the beginning point being a point which is closest to the projected first threshold magnitude on a first side of the endpoint of MV; and projecting the second threshold magnitude around the endpoint of MV to determine the ending point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the ending point as a second sample point of the transformed ECG signal, the ending point being a point which is closest to the projected second threshold magnitude on a second side of the endpoint of MV.
  • 41. The method of claim 39, wherein: projecting the first threshold magnitude around the endpoint of MV to determine the beginning point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the beginning point as an interpolated beginning point between a first sample point and a second sample point of the transformed ECG signal on a first side of the endpoint of MV, the first and second sample points being the two closest points to the projected first threshold magnitude on the first side of the endpoint of MV; and projecting the second threshold magnitude around the endpoint of MV to determine the ending point on the transformed ECG signal in the plane defined by ev1 and ev2 comprises determining the ending point as an interpolated ending point between a third sample point and a fourth sample point of the transformed ECG signal on a second side of the endpoint of MV, the third and fourth sample points being the two closest points to the projected second threshold magnitude on the second side of the endpoint of MV.
  • 42. The method of claim 39, wherein determining the repolarization duration which is based on the maximum vector MV comprises: determining a beginning time corresponding to the beginning point; determining a peak time corresponding to the endpoint of MV; and determining an early repolarization duration (ERD) for the first threshold percentage which comprises a time difference between the peak time and the beginning time.
  • 43. The method of claim 42, wherein determining the repolarization duration which is based on the maximum vector MV further comprises: determining an ending time corresponding to the ending point; and determining a late repolarization duration (LRD) for the second threshold percentage which comprises a time difference between the ending time and the peak time.
  • 44. The method of claim 43, wherein determining the repolarization duration which is based on the maximum vector MV further comprises: determining a total repolarization duration (TRD) comprising the sum of ERD and LRD.
  • 45. A computer readable medium having stored thereon instructions for assessing repolarization abnormalities, which, when executed by a processor, causes the processor to perform the steps according to claim 1.
  • 46. A system for assessing repolarization abnormalities, comprising: a processor configured to determine a repolarization duration which is based on a maximum vector MV from transformed ECG repolarization signals in a plane defined by eigenvectors ev1 and ev2 which result from principal component analysis on the ECG repolarization signals from ECG data; a data input coupled to the processor and configured to provide the processor with the ECG data; and a user interface coupled to either the processor or the data input.
  • 47. The system of claim 46, wherein the processor is configured to be at least partially implantable in a subject's body.
  • 48. The system of claim 46, further comprising a database coupled to the processor.
  • 49. The system of claim 46, further comprising a database coupled to the data input.
  • 50. The system of claim 46, further comprising an ECG capture device coupled to the data input.
  • 51. The system of claim 50, wherein the ECG capture device is selected from the group consisting of a Holter monitor; a twelve-lead monitor; an 8 lead monitor; a monitor using a bipolar lead system, and a monitor using a unipolar lead system.
  • 52. The system of claim 50, wherein the ECG capture device is coupled to the data input by a wireless connection.
  • 53. The system of claim 52, wherein the ECG capture device is configured to transmit data to the data input during times which do not interfere with data measurement times of the ECG capture device.
  • 54. The system of claim 50, wherein at least a portion of the ECG capture device is implantable in a subject's body.
  • 55. The system of claim 54, further comprising a treatment device coupled to the processor, and wherein the processor is further configured to activate the treatment device to attempt to correct a repolarization abnormality indicated by the repolarization duration.
  • 56. The system of claim 55, wherein the treatment device is implantable in the subject's body.
  • 57. The system of claim 55, wherein the treatment device comprises a pharmacological agent administrator.
  • 58. The system of claim 55, wherein the treatment device comprises a defibrillator.
  • 59. The system of claim 46, wherein the processor and the data input are coupled together via a network.
  • 60. The system of claim 46, further comprising: an ECG capture device coupled to the data input; and a pharmacological agent administrator coupled to the processor.
  • 61. The system of claim 60, wherein the processor is further configured to: administer a pharmacological agent with the pharmacological agent administrator; capture a post-administration set of ECG signals; and determine a post-administration repolarization duration based on at least two ECG signals in the post-administration set of ECG signals.
  • 62. The system of claim 61, wherein the processor is further configured to: prior to administering the pharmacological agent, capture a pre-administration set of ECG signals; and determine a pre-administration repolarization duration based on at least two ECG signals in the pre-administration set of ECG signals.
  • 63. The system of claim 62, wherein the processor is further configured to compare the pre-administration repolarization duration with the post-administration repolarization duration.
  • 64. The system of claim 46, wherein the repolarization duration comprises an early repolarization duration (ERD) determined by: taking a first threshold percentage of a magnitude of MV to determine a first threshold magnitude; projecting the first threshold magnitude relative to an endpoint of MV to determine a beginning point on the transformed ECG signal prior to the endpoint of MV; determining a beginning time corresponding to the beginning point; determining a peak time corresponding to the end point of MV; and determining the ERD for the first threshold percentage, the ERD comprising a time difference between the peak time and the beginning time.
  • 65. The system of claim 64, wherein the repolarization duration also comprises a late repolarization duration (LRD) further determined by: taking a second threshold percentage of the magnitude of MV to determine a second threshold magnitude; projecting the second threshold magnitude relative to the endpoint of MV to determine an ending point on the transformed ECG signal after the endpoint of MV; determining an ending time corresponding to the ending point; and determining the LRD for the second threshold percentage, the LRD comprising a time difference between the ending time and the peak time.
  • 66. The system of claim 65, wherein the repolarization duration further comprises a total repolarization duration (TRD) which comprises the sum of the ERD and the LRD.
  • 67. The system of claim 65, wherein the first threshold percentage equals the second threshold percentage.
  • 68. The system of claim 46, wherein the user interface is configured to display the ECG data.
  • 69. The system of claim 46, wherein the user interface is configured to display a T-loop plotted in a preferential plane.
  • 70. The system of claim 46, wherein the user interface is configured to display a primary component the ECG repolarization signals versus time.
  • 71. The system of claim 46, wherein the user interface is configured to display a calculated parameter selected from the group consisting of ERD, LRD, and TRD, the calculated parameter being determined for at least one threshold percentage.
  • 72. The system of claim 46, wherein the user interface is configured to allow a user to select the ECG data from a database.
  • 73. The system of claim 46, wherein the user interface is configured to allow a user to start and stop collecting data from an ECG capture device.
  • 74. A method for analyzing an effect of a pharmacological agent on heart repolarization, comprising: obtaining a first set of ECG signals; identifying at least two repolarization signals from the first set of ECG signals; principal component analysis on the at least two repolarization signals from the first set of ECG signals to extract at least first eigenvectors ev1-1 and ev1-2; determining a first maximum vector MV1 based on a first transformed ECG signal in a first plane defined by ev1-1 and ev1-2; determining a first repolarization duration which is based on the first maximum vector MV1; administering the pharmacological agent; obtaining a second set of ECG signals; identifying at least two repolarization signals from the second set of ECG signals; principal component analysis on the at least two repolarization signals from the second set of ECG signals to extract at least second eigenvectors ev2-1 and ev2-2; determining a second maximum vector MV2 based on a second transformed ECG signal in a second plane defined by ev2-1 and ev2-2; determining a second repolarization duration which is based on the second maximum vector MV2; and detecting a repolarization abnormality based on the first repolarization duration and the second repolarization duration.
  • 75. A method for assessing repolarization abnormalities, comprising: a) low-pass FIR filtering a set of at least two ECG signals; b) discarding one or more leading beats from the ECG signals; c) discarding one or more trailing beats from the ECG signals; d) discarding beats from the ECG signals which have a heart rate that varies by more than ten percent in beats of the previous two minutes; e) identifying at least two repolarization signals from the at least two ECG signals by defining each of the at least two repolarization signals as starting from a point J and ending at a point which is approximately 220 milliseconds prior to a following R peak; f) principal component analysis on the at least two repolarization signals to extract at least eigenvectors ev1 and ev2, wherein the principal component analysis comprises: 1) subtracting a mean from the at least two repolarization signals to produce a data set with a mean of zero; 2) calculating a covariance matrix for the data set; 3) calculating a set of eigenvectors and corresponding eigenvalues for the covariance matrix; 4) determining that a first eigenvector from the set of eigenvectors with the largest corresponding eigenvalue is the eigenvector ev1; and 5) determining that a second eigenvector from the set of eigenvectors with the second-largest corresponding eigenvalue is the eigenvector ev2; g) determining the maximum vector MV based on a transformed ECG signal in a plane defined by ev1 and ev2; such that MV=max{right arrow over (VECG)}(t)−{right arrow over (VECG)}(TQ) where {right arrow over (VECG)}(t)={right arrow over (ev)}1(t)+{right arrow over (ev)}2(t) and TQ is a time coinciding with a beginning of QRS complex for the transformed ECG signal; h) determining an early repolarization duration (ERD) which is based on the maximum vector MV at a threshold percentage x % of MV, such that: ERDx %=TMV−TE, where TE is a value for t where the following equation is fulfilled: ∥VECG(t)−VECG(TMV)∥=MV·x %, with tTMV and j) determining a total repolarization duration (TRD) which is based on the maximum vector MV at the threshold percentage x % of MV, such that: TRDx %=ERDx %+LRDx %.