Automatic reduction of stimulus polarization artifact for accurate evaluation of ventricular evoked responses

The ventricular evoked response, the cardiac depolarization generated in response to a pacing stimulus, is potentially useful as a sensor for rate responsive pacing and automatic threshold tracking. It is necessary to minimize the polarization artifact that results from pacing in order to sense cardiac depolarizations from the same electrodes that pace the heart. To accomplish this, a triphasic stimulus waveform consisting of precharge, stimulus, and postcharge was used. An algorithm was developed that introduced pacing stimuli during the refractory period of sensed beats, when cardiac depolarization could not occur by definition and polarization artifact could be evaluated. Precharge duration was varied until the amplitude of the polarization artifact was small compared to the evoked response. In 18 patients with temporary electrode catheters, polarization artifact was reduced from 6.8 +/- 3.4 mV to 1.9 +/- 1.1 mV after balancing (P less than 0.005). Initial precharge duration was 3200 mu sec and the mean final precharge duration was 3551 +/- 516 mu sec. In 14 patients with permanent bipolar pacing leads, polarization artifact was reduced from 3.2 +/- 3.5 mV to 0.7 +/- 0.6 mV (P less than 0.025). Final precharge duration averaged 3440 +/- 310 mu sec. Under a wide variety of pacing conditions, this algorithm simply and quickly reduces polarization artifact to a minimum to allow accurate analysis of evoked responses.