Multi-channel cardiac measurements

1. An automatic method of determining local activation time (LAT) in multi-channel cardiac electrogram signals including a plurality of cardiac channels, the method comprising:

storing the cardiac channel signals;

selecting a mapping channel, a ventricular channel, and a reference channel from among the plurality of cardiac channels;

using the selected channels to compute first LAT values at a plurality of mapping-channel locations;

monitoring the quality of at least one selected channel;

if the quality of a monitored cardiac channel falls below a standard, replacing the sub-standard channel with another channel of the plurality of channels having an above-standard quality; and

computing second LAT values based on the replacement cardiac channel.

2. The automatic LAT-determination method of claim 1 wherein computing second LAT values based on the replacement cardiac channel includes computing second LAT values for previous mapping-channel locations.

3. The automatic LAT-determination method of claim 1 further including displaying the first LAT values as a first LAT map, and if second LAT values are computed, generating a second LAT map based on the second LAT values.

4. The automatic LAT-determination method of claim 1 wherein the replaced and replacement channels are reference channels.

5. The automatic LAT-determination method of claim 4 wherein monitoring the signal quality of a cardiac channel includes:

generating a velocity-dependent signal from the monitored channel signal over a preset time window;

finding the median value MED of the velocity-dependent signal over the preset time window;

dividing the velocity-dependent signal into a plurality of signal chunks;

finding the chunk-maximum within each chunk;

finding the maximum MAX and minimum MIN values among the plurality of chunk-maxima; and

computing a channel signal quality SQ based on MAX, MIN and MED.

6. The automatic LAT-determination method of claim 5 wherein the channel signal quality SQ is computed as 2×MIN−MAX−K 1 ×MED where K 1 is a constant.

7. The automatic LAT-determination method of claim 6 wherein each signal chunk is of equal length.

8. The automatic LAT-determination method of claim 7 wherein each velocity-dependent signal is divided into three equal-length chunks.

9. The automatic LAT-determination method of claim 8 wherein computing second LAT values based on the replacement cardiac channel includes computing second LAT values for previous mapping-channel locations.

10. The automatic LAT-determination method of claim 8 further including displaying the first LAT values as a first LAT map, and if second LAT values are computed, generating a second LAT map based on the second LAT values.

11. The automatic LAT-determination method of claim 5 wherein the preset time window is about 6 seconds long.

12. The automatic LAT-determination method of claim 11 wherein the channel signal quality is computed as 2×MIN−MAX−K 1 ×MED where K 1 is a constant.

13. The automatic LAT-determination method of claim 12 wherein K 1 is about 4.

14. The automatic LAT-determination method of claim 13 wherein each signal chunk is of equal length.

15. The automatic LAT-determination method of claim 14 wherein each velocity-dependent signal is divided into three equal-length chunks.

16. The automatic LAT-determination method of claim 15 wherein computing second LAT values based on the replacement cardiac channel includes computing second LAT values for previous mapping-channel locations.

17. The automatic LAT-determination method of claim 15 further including displaying the first LAT values as a first LAT map, and if second LAT values are computed, generating a second LAT map based on the second LAT values.

18. The automatic LAT-determination method of claim 1 wherein monitoring the quality of cardiac channels includes monitoring cycle-length variability of each channel of a first subset of the monitored channels.

19. The automatic LAT-determination method of claim 18 wherein the first-subset cycle-length variability monitoring includes determining a maximum cycle length MAX CL , a minimum cycle length MIN CL , and a cycle-length variability VAR RC for each monitored first-subset channel.

20. The automatic LAT-determination method of claim 19 wherein determining a maximum cycle length MAX CL , a minimum cycle length MIN CL , and a cycle-length variability VAR RC includes for each monitored first-subset channel:

selecting a set of sequential sub-signals;

estimating cycle lengths and cycle-length variability for each sub-signal;

selecting MAX CL and MIN CL as the maximum and minimum of the sub-signal cycle lengths; and selecting VAR RC as the maximum sub-signal variability.

21. The automatic LAT-determination method of claim 20 wherein estimating the cycle length of each sub-signal includes:

estimating pulse intervals in the ventricular channel;

generating a velocity-dependent signal from the sub-signal over a preset time window;

autocorrelating the sub-signal velocity-dependent signal;

selecting a peak value of the autocorrelation based on ventricular pulse-interval estimates; and

setting the sub-signal cycle length CL to the lag value of the selected peak in the autocorrelation.

22. The automatic LAT-determination method of claim 21 wherein estimating sub-signal cycle-length variability includes:

estimating the lag DCL at an autocorrelation peak near twice CL;

computing an intermediate cycle-length CL A ; and

computing sub-signal cycle-length variability as the absolute value of the difference between CL and CL A .

23. The automatic LAT-determination method of claim 20 wherein the sub-signals are of equal length.

24. The automatic LAT-determination method of claim 23 wherein the sub-signals are about 6 seconds long.

25. The automatic LAT-determination method of claim 24 wherein the set of sub-signals comprises 5 sub-signals.

26. The automatic LAT-determination method of claim 25 wherein estimating the cycle length of each sub-signal includes:

estimating pulse intervals in the ventricular channel;

generating a velocity-dependent signal from the sub-signal over a preset time window;

autocorrelating the sub-signal velocity-dependent signal;

selecting a peak value of the autocorrelation based on ventricular pulse-interval estimates; and

setting the sub-signal cycle length CL to the lag value of the selected peak in the autocorrelation.

27. The automatic LAT-determination method of claim 26 wherein estimating sub-signal cycle-length variability includes:

estimating the lag DCL at an autocorrelation peak near twice CL;

computing an intermediate cycle-length CL A ; and

computing sub-signal cycle-length variability as the absolute value of the difference between CL and CL A .

28. The automatic LAT-determination method of claim 20 wherein monitoring the quality of cardiac channels further includes for each first-subset channel:

determining a signal quality value SQ i for each sub-signal;

summing the sub-signal signal quality values SQ i to generate a first-subset channel signal quality value SQ RC ; and

computing a first-subset channel figure-of-merit FM RC .

29. The automatic LAT-determination method of claim 28 wherein the figure-of-merit FM RC is computed as SQ RC /K 2 −MAX CL −MIN CL −K 3 ×VAR RC where K 2 and K 3 are constants.

30. The automatic LAT-determination method of claim 29 wherein the of K 2 is about 32 and the value of K 3 is about 2.

31. The automatic LAT-determination method of claim 29 wherein if the quality of a monitored first-subset channel falls below a standard, replacing the sub-standard first-subset channel with another first-subset channel having a higher figure-of-merit.

32. The automatic LAT-determination method of claim 31 wherein the of K 2 is about 32 and the value of K 3 is about 2.

33. The automatic LAT-determination method of claim 1 wherein monitoring the quality of cardiac channels includes monitoring pulse-interval variability of each channel of a second subset of the monitored channels.

34. The automatic LAT-determination method of claim 33 wherein the second-subset pulse-interval variability monitoring includes for each monitored second-subset channel:

selecting a set of sequential sub-signals;

for each sub-signal, determining a maximum pulse interval MAX Ei and a minimum pulse-interval MIN Ei ;

for each sub-signal, computing a sub-signal pulse-interval variability VAR i as MAX Ei −MIN Ei ; and

computing a second-subset channel pulse-interval variability VAR VC as the maximum value among the sub-signal pulse-interval variability values VAR i .

35. The automatic LAT-determination method of claim 34 wherein the sub-signals are of equal length.

36. The automatic LAT-determination method of claim 35 wherein the sub-signals are about 6 seconds long.

37. The automatic LAT-determination method of claim 34 wherein the set of sub-signals comprises 5 sub-signals.

38. The automatic LAT-determination method of claim 34 wherein monitoring the quality of cardiac channels further includes for each second-subset channel:

determining a signal quality value SQ i for each sub-signal; and

summing the sub-signal signal quality values SQ i to generate a second-subset channel signal quality value SQ VC .

39. The automatic LAT-determination method of claim 34 wherein if the quality of a monitored second-subset channel falls below a standard, replacing the sub-standard second-subset channel with another second-subset channel having a higher channel signal quality SQ VC .

40. The automatic LAT-determination method of claim 39 wherein the channel pulse-interval variability VAR VC of the another second-subset channel is below a computed threshold pulse-interval variability T VC .

41. The automatic LAT-determination method of claim 40 wherein threshold T VC is proportional to the median of all second-subset channel pulse-interval variability values VAR VC .

42. The automatic LAT-determination method of claim 18 wherein monitoring the quality of cardiac channels further includes monitoring pulse-interval variability of each channel of a second subset of the monitored channels.

43. The automatic LAT-determination method of claim 1 wherein monitoring the quality of a cardiac channel includes estimating an LAT measurement confidence interval and comparing the estimate with a measurement-confidence criterion.

44. The automatic LAT-determination method of claim 43 wherein estimating an LAT measurement confidence interval includes:

estimating the cycle-length CL of the reference channel;

selecting an activation at a time t M in the mapping-channel signal;

identifying a plurality of activations at times t R in the reference channel;

computing a plurality of LAT values with respect to the values of t R ;

selecting a subset of the plurality of LAT values; and

estimating the LAT measurement confidence interval as the difference between the maximum and minimum LAT values in the subset.

45. The automatic measuring method of claim 44 wherein the subset comprises the interquartile values of the plurality of LAT values.

46. The automatic LAT-determination method of claim 1 wherein monitoring the quality of a cardiac channel includes estimating a channel cycle-length confidence interval.

47. The automatic LAT-determination method of claim 46 wherein estimating a channel cycle-length confidence interval includes estimating channel cycle-length by:

estimating pulse intervals in the ventricular channel;

generating a velocity-dependent signal from the channel signal over a preset time window;

autocorrelating the channel velocity-dependent signal;

selecting a peak value of the autocorrelation based on ventricular pulse-interval estimates; and

setting the channel cycle length CL to the lag value of the selected peak in the autocorrelation.

48. The automatic LAT-determination method of claim 47 wherein estimating a channel cycle-length confidence interval includes:

determining the lag value L2 at a peak in the autocorrelation near twice CL; and

estimating the channel cycle-length confidence interval as a function of the difference between L2 and twice CL.

49. The automatic LAT-determination method of claim 1 wherein monitoring the quality of a cardiac channel includes estimating a channel pulse-interval confidence interval.

50. The automatic LAT-determination method of claim 49 wherein estimating a channel pulse-interval confidence interval includes estimating channel pulse-interval by:

generating a velocity-dependent signal from the channel signal over a preset time window;

identifying a plurality of activation times in the channel velocity-dependent signal;

computing a set of activation intervals;

identifying maximum and minimum activation intervals in the set; and

setting the confidence interval equal to the difference between the maximum and minimum activation intervals.