Methods and apparatuses for cardiac arrhythmia classification using dynamic beatdriven morphological feature extraction

Abstract

An implantable cardioverter/defibrillator (ICD) includes a tachyarrhythmia detection and classification system that classifies tachyarrhythmias based on a morphological analysis of arrhythmic waveforms and a template waveform. Correlation coefficients each computed between morphological features of an arrhythmic waveform and morphological features of the template waveform provide for the basis for classifying the tachyarrhythmia. In one embodiment, morphological features are collected from a sensed arrhythmic waveform, and temporally corresponding morphological features are extracted a stored template waveform.

1 . A system for classifying tachyarrhythmias, the system comprising:
 a template generation circuit configured to produce a template waveform associated with a template heart beat of a known type cardiac rhythm;   a template waveform storage circuit coupled to the template generation circuit, the template waveform storage circuit configured to store the template waveform;   a feature locating circuit configured to select a plurality of arrhythmic morphological features on an arrhythmic waveform associated with an arrhythmic heart beat of a tachyarrhythmia and produce timing information indicative of locations of the plurality of arrhythmic morphological features on the arrhythmic waveform; and   a feature extracting circuit coupled to the template waveform storage circuit and the feature locating circuit, the feature extracting circuit configured to locate a plurality of template morphological features on the stored template waveform using the timing information.   
     
     
         2 . The system of  claim 1 , wherein the template generation circuit comprises:
 a template waveform alignment circuit configured to align a plurality of waveforms each associated with a heart beat of the known type cardiac rhythm by a fiducial point; and   a template averaging circuit coupled to the template waveform alignment circuit, the template averaging circuit configured to produce the template waveform by averaging amplitudes of the plurality of waveforms.   
     
     
         3 . The system of  claim 2 , wherein the template waveform alignment circuit comprises a peak detector configured to detect a peak of a depolarization, and wherein the fiducial point is the peak of the depolarization. 
     
     
         4 . The system of  claim 1 , wherein the feature locating circuit comprises:
 a waveform input configured to receive the arrhythmic waveform;   a fiducial point locating circuit coupled to the waveform input, the fiducial point locating circuit configured to detect a fiducial point on the arrhythmic waveform;   a feature selecting circuit coupled to the fiducial point locating circuit, the feature selecting circuit configured to select the plurality of arrhythmic morphological features on the arrhythmic waveform based on one or more predetermined criteria; and   a feature timing circuit coupled to the feature selecting circuit, the feature timing circuit configured to measure time intervals each between the fiducial point on the arrhythmic waveform and one arrhythmic morphological feature of the selected plurality of arrhythmic morphological features.   
     
     
         5 . The system of  claim 4 , wherein the fiducial point locating circuit comprises a peak detector configured to detect a peak of a depolarization on the arrhythmic waveform. 
     
     
         6 . The system of  claim 4 , wherein the feature extracting circuit is configured to locate the plurality of template morphological features on the template waveform based on the fiducial point on the arrhythmic waveform and the measured time intervals. 
     
     
         7 . The system of  claim 6 , wherein the feature extracting circuit is configured to locate a corresponding fiducial point on the template waveform, align the fiducial point on the arrhythmic waveform and the corresponding fiducial point on the template waveform, and locate the plurality of template morphological features on the template waveform using the measured time intervals each as a time interval between one template morphological feature of the plurality of template morphological features and the fiducial point on the template waveform. 
     
     
         8 . The system of  claim 1 , further comprising a correlation analyzer configured to produce a correlation coefficient indicative of a correlation between the selected plurality of arrhythmic morphological features and the located plurality of template morphological features. 
     
     
         9 . The system of  claim 8 , wherein the tachyarrhythmia comprises a 1:1 tachyarrhythmia, and further comprising a 1:1 tachyarrhythmia classifier configured to classify the arrhythmic heart beat using the correlation coefficient and a predetermined correlation threshold. 
     
     
         10 . The system of  claim 9 , wherein the 1:1 tachyarrhythmia classifier is configured to classify the arrhythmic heart beat as about of a first type tachyarrhythmia response to the correlation coefficient being greater than the predetermined correlation threshold and a beat of a second type tachyarrhythmia in response the correlation coefficient not being greater than the predetermined correlation threshold. 
     
     
         11 . A method for classifying tachyarrhythmias, the method comprising:
 producing a template waveform associated with a template heart beat;   storing the template waveform;   receiving an arrhythmic waveform associated with an arrhythmic heart beat;   selecting a plurality of arrhythmic morphological features on the arrhythmic waveform;   producing timing information indicative of locations of the plurality of arrhythmic morphological features on the arrhythmic waveform; and   locating a plurality of template morphological features on the template waveform using the timing information.   
     
     
         12 . The method of  claim 11 , wherein receiving the arrhythmic waveform comprises receiving a waveform associated with a heart beat sensed during a 1:1 tachyarrhythmia. 
     
     
         13 . The method of  claim 12 , further comprising:
 producing a correlation coefficient indicative of a correlation between the selected plurality of arrhythmic morphological features and the located plurality of template morphological features; and   classifying the arrhythmic heart beat using the correlation coefficient and a predetermined correlation threshold.   
     
     
         14 . The method of  claim 13 , further comprising classifying the arrhythmic heart beat as a beat of a first type tachyarrhythmia in response to the correlation coefficient being greater than the predetermined correlation threshold and a beat of a second type tachyarrhythmia in response to the correlation coefficient not being greater than the predetermined correlation threshold. 
     
     
         15 . The method of  claim 14 , wherein the template heart beat is a heart beat sensed during a normal sinus rhythm, the first type tachyarrhythmia is a supraventricular tachyarrhythmia, and the second type tachyarrhythmia is a ventricular tachyarrhythmia. 
     
     
         16 . The method of  claim 13 , wherein producing the template waveform comprises:
 aligning a plurality of waveforms each representing a heart beat of a known type cardiac rhythm by a fiducial point; and   producing the template waveform by averaging amplitudes of the plurality of waveforms each representing the heart beat of the known type cardiac rhythm.   
     
     
         17 . The method of  claim 13 , wherein producing the timing information indicative of the locations of the plurality of arrhythmic morphological features on the arrhythmic waveform comprises:
 detecting a fiducial point on the arrhythmic waveform; and   measuring time intervals each between the fiducial point and one arrhythmic morphological feature of the selected plurality of arrhythmic morphological features on the arrhythmic waveform.   
     
     
         18 . The method of  claim 17 , wherein detecting the fiducial point on the arrhythmic waveform comprises detecting a peak of a depolarization on the arrhythmic waveform. 
     
     
         19 . The method of  claim 17 , wherein locating the plurality of template morphological features on the template waveform comprises locating the plurality of template morphological features on the template waveform based on the fiducial point on the arrhythmic waveform and the measured time intervals. 
     
     
         20 . The method of  claim 19 , wherein locating the plurality of template morphological features on the template waveform comprises:
 locating a corresponding fiducial point on the template waveform;   aligning the fiducial point on the arrhythmic waveform and the corresponding fiducial point on the template waveform; and   locating the plurality of template morphological features on the template waveform using the measured time intervals each as a time interval between one template morphological feature of the plurality of template morphological features and the fiducial point on the template waveform.