Introduction: Fractionated, low-amplitude or long-duration electrograms have limited specificity for locating reentrant circuits causing ventricular tachycardia (VT). In this study a new method is described, adaptive template matching (ATM), based on the quantification of beat-to-beat changes in electrograms, for locating functional reentrant circuits that are relatively stable and cause monomorphic VT.
Methods and results: Monomorphic VTs were induced in 4-day-old infarcted canine hearts by programmed stimulation and reentrant circuits mapped in the epicardial border zone with a 196 or 312 bipolar electrode array. For ATM analysis, a template electrogram from each electrode, during an early cycle, was matched with all subsequent (input) electrograms at the same site by weighting the inputs of amplitude, duration, average baseline, and phase lag. The mean square error (MSE) between template and input was the criterion used to adapt the weights, and was also a measure of changes in electrogram shape that occur from cycle to cycle. The variance of each of the weighting parameters at all electrode sites were plotted on a representation of the electrode array, and the location of the functional lines of block bounding the central common pathway of reentrant circuits with figure-of-eight characteristics, overlaid on the ATM map. Peaks of high variance were found to be coincident with functional lines of block during all tachycardia episodes.
Conclusion: Specific beat-to-beat changes in electrograms occur at functional lines of block in reentrant circuits that can be quantified by ATM analysis, suggesting that these regions might be located without activation mapping. The method might be useful to guide ablation catheter position.