Beat-morphology matching scheme for cardiac sensing and event detection

1. A method for classifying cardiac beats for use in detecting cardiac rhythms, comprising:

sensing a cardiac signal over a plurality of cardiac cycles;

determining a template of a known cardiac signal in response to the cardiac signal sensed over the plurality of cardiac cycles;

determining a template alignment point from the template of the known cardiac signal;

sensing a cardiac signal having an unknown morphology over a plurality cardiac cycles;

determining a fourth order difference signal by applying a fourth order difference function to the cardiac signal having an unknown morphology;

determining an alignment point for the cardiac signal having an unknown morphology from the fourth order difference signal;

aligning the template and the cardiac signal having an unknown morphology across an alignment window by aligning the template alignment point and the alignment point of the cardiac signal having an unknown morphology, resulting in an aligned fourth order difference signal of the cardiac signal having an unknown morphology;

determining an R-wave onset and an R-wave offset from the aligned fourth order difference signal of the cardiac signal having an unknown morphology;

determining an R-wave width as a difference between the R-wave onset and the R-wave offset;

determining a morphology analysis window in response to the R-wave width;

determining a first morphology match metric between the aligned template and the cardiac signal having an unknown morphology across the morphology analysis window; and

classifying the cardiac signal having an unknown morphology based on the first morphology match metric.

2. The method of claim 1 , wherein the morphology analysis window comprises a first number of sample points of the cardiac signal having an unknown morphology, and the alignment window comprises a second number of sample points of the cardiac signal having an unknown morphology, the second number greater than the first number.

3. The method of claim 1 , wherein determining the template and determining the template alignment point comprise:

determining a fourth order difference signal from the cardiac signal over each of the plurality of cardiac cycles;

determining an alignment point from the fourth order difference signal of each of the plurality of cardiac cycles;

aligning the plurality of cardiac cycle signals by aligning the alignment point in each of the plurality of cardiac cycles; and

averaging the aligned plurality of cardiac cycle signals.

4. The method of claim 1 , wherein determining the template and determining the template alignment point comprise:

computing, with a processor, a fourth order difference signal from the cardiac signal over each of the plurality of cardiac cycles;

determining an alignment point from the fourth order difference signal of each of the plurality of cardiac cycles;

aligning the plurality of fourth order difference signals of each of the cardiac cycle signals by aligning the alignment points; and

averaging the aligned plurality of fourth order difference signals.

5. The method of claim 1 , wherein aligning the template alignment point and the cardiac signal having an unknown morphology comprises determining a peak amplitude and a polarity of a dominant pulse of a fourth order difference signal obtained from the template.

6. The method of claim 5 , wherein determining the alignment point for the cardiac signal having an unknown morphology comprises:

determining a peak amplitude of a dominant pulse of the fourth order difference signal obtained from the cardiac signal having an unknown morphology that matches the polarity of the template alignment point.

7. The method of claim 1 , further comprising:

after aligning the template alignment point and the alignment point of the cardiac signal having an unknown morphology, shifting the template by shifting the template alignment point relative to the alignment point of the cardiac signal having an unknown morphology by at least one sample point;

determining a second morphology match metric as a measure of a similarity between the shifted template and the cardiac signal having an unknown morphology;

selecting a greatest one of the first morphology match metric and the second morphology match metric; and

classifying the cardiac signal having an unknown morphology in response to the selected one of the first and second morphology match metrics.

8. The method of claim 1 , further comprising:

after aligning the template alignment point and the alignment point of the cardiac signal having an unknown morphology, shifting the template a plurality of times by shifting the template alignment point relative to the alignment point of the cardiac signal having an unknown morphology by at least one sample point to obtain a plurality of alignments between the template and the cardiac signal cardiac signal having an unknown morphology;

determining a morphology match score for each one of the plurality of alignments;

selecting a greatest one of the first morphology match metric and the morphology match metrics for the plurality of alignments; and

classifying the cardiac signal having an unknown morphology in response to the selected one of the first morphology match metric and the morphology match metrics for the plurality of alignments.

9. The method of claim 1 , wherein determining the morphology match metric comprises computing a normalized waveform area difference between the aligned template and the cardiac signal having an unknown morphology.

10. A medical device for classifying cardiac beats for use in detecting cardiac rhythms, comprising:

a plurality of electrodes for sensing a cardiac signal over a plurality of cardiac cycles; and

a processor configured to:

determine a template of a known cardiac signal in response to the cardiac signal sensed over the plurality of cardiac cycles,

determine a template alignment point from the template of the known cardiac signal,

sense an cardiac signal having an unknown morphology over a plurality of cardiac cycles,

determine a fourth order difference signal by applying a fourth order difference function to the cardiac signal having an unknown morphology,

determine an alignment point for the cardiac signal having an unknown morphology from the fourth order difference signal,

align the template and the cardiac signal having an unknown morphology across an alignment window by aligning the template alignment point and the alignment point of the cardiac signal having an unknown morphology, resulting in an aligned fourth order difference signal of the cardiac signal having an unknown morphology,

determine an R-wave onset and an R-wave offset from the fourth order difference signal of the cardiac signal having an unknown morphology,

determine an R-wave width as a difference between the R-wave onset and the R-wave offset,

determine a morphology analysis window in response to the R-wave width,

determine a first morphology match metric between the aligned template and the cardiac signal having an unknown morphology across the morphology analysis window, and

classify the cardiac signal having an unknown morphology based on the first morphology match metric.

11. The medical device of claim 10 , wherein the morphology analysis window comprises a first number of sample points of the cardiac signal having an unknown morphology, and the alignment window comprises a second number of sample points of the cardiac signal having an unknown morphology, the second number greater than the first number.

12. The medical device of claim 10 , wherein determining the template and the template alignment point comprise:

determining a fourth order difference signal in response to the cardiac signal over each of the plurality of cardiac cycles;

determining an alignment point from the fourth order difference signal of each of the plurality of cardiac cycles;

aligning the plurality of cardiac cycle signals by aligning the alignment point in each of the plurality of cardiac cycles; and

averaging the aligned plurality of cardiac cycle signals.

13. The medical device of claim 10 , wherein determining the template and the template alignment point, wherein the processor is further configured to:

compute a fourth order difference signal from the cardiac signal over each of the plurality of cardiac cycles;

determine an alignment point from the fourth order difference signal of each of the plurality of cardiac cycles;

align the plurality of fourth order difference signals of each of the cardiac cycle signals by aligning the alignment points; and

averaging the aligned plurality of fourth order difference signals.

14. The medical device of claim 10 , wherein aligning the template alignment point and the cardiac signal having an unknown morphology comprises determining a peak amplitude and a polarity of a dominant pulse of a fourth order difference signal corresponding to the template.

15. The medical device of claim 14 , wherein determining the alignment point of the cardiac signal having an unknown morphology comprises:

determining a peak amplitude of a dominant pulse of the fourth order difference signal obtained from the cardiac signal having an unknown morphology that matches the polarity of the template alignment point.

16. The medical device of claim 10 , wherein the processor is configured to:

after aligning the template alignment point and the alignment point of the cardiac signal having an unknown morphology, shift the template by shifting the template alignment point relative to the alignment point of the cardiac signal having an unknown morphology by at least one sample point,

determine a second morphology match metric as a measure of a similarity between the shifted template and the cardiac signal having an unknown morphology,

select a greatest one of the first morphology match metric and the second morphology match metric, and

classify the cardiac signal having an unknown morphology in response to the selected one of the first and second morphology match metrics.

17. The medical device of claim 10 , wherein the processor is configured to

after aligning the template alignment point and the alignment point of the cardiac signal having an unknown morphology, shift the template a plurality of times by shifting the template alignment point relative to the alignment point of the cardiac signal having an unknown morphology by at least one sample point to obtain a plurality of alignments between the template and the cardiac signal having an unknown morphology,

determine a morphology match score for each one of the plurality of alignments, select a greatest one of the first morphology match metric and the morphology match metrics for the plurality of alignments, and

classify the cardiac signal having an unknown morphology in response to the selected one of the first morphology match metric and the morphology match metrics for the plurality of alignments.

18. The medical device of claim 10 , wherein determining the morphology match metric comprises computing a normalized waveform area difference between the aligned template and the cardiac signal having an unknown morphology.

19. A non-transitory, computer-readable medium storing a set of instructions, which when executed by a processor of a medical device causes the device to:

sense a cardiac signal over a plurality of cardiac cycles using a plurality of electrodes coupled to a sensing module;

determine a template of a known cardiac signal in response to the cardiac signal sensed over the plurality of cardiac cycles;

determine a template alignment point from the template of the known cardiac signal;

sense an cardiac signal having an unknown morphology over a plurality of cardiac cycles;

determine a fourth order difference signal by applying a fourth order difference function to the cardiac signal having an unknown morphology;

determine an alignment point for the cardiac signal having an unknown morphology from the fourth order difference signal;

align the template and the cardiac signal having an unknown morphology across an alignment window by aligning the template alignment point and the alignment point of the cardiac signal having an unknown morphology, resulting in an aligned fourth order difference signal of the cardiac signal having an unknown morphology;

determine an R-wave onset and an R-wave offset in response to the fourth order difference signal of the cardiac signal having an unknown morphology;

determine an R-wave width as a difference between the R-wave onset and the R-wave offset;

determine a morphology analysis window in response to the R-wave width;

determine a first morphology match metric between the aligned template and the cardiac signal having an unknown morphology across the morphology analysis window; and

classify the cardiac signal having an unknown morphology based on the first morphology match metric.