Methods
Studies conformed to the guiding principles of the Declaration of Helsinki. In 15 dogs (14-25 kg) anesthesia was induced and maintained by Nembutal 30 mg/kg. Positive pressure ventilation was maintained through an endocardial tube to maintain normal blood gases. The chest was opened through a medial sternotomy. The left anterior descending coronary artery was dissected free along a 5 mm segment proximal to the first diagonal branch. The left anterior descending coronary artery was abruptly occluded with a hemoclip for a period of 15 minutes. 768 unipolar electrodes on 112 transmural multipolar needle electrodes were inserted into the heart in a standardized array illustrated in Figure 1. Care was taken to avoid epicardial vessels. Each needle was fabricated from 23 gauge steel with either 4 or 8 tungsten wire electrodes set in epoxy at 1, 2 or 4 mm intervals (8). This array provided a total of 768 unipolar electrograms with a reference electrode placed adjacent to the left first costosternal junction. A 20 to 30 minute control period was allowed after insertion of the last needle for injury currents to subside. At the conclusion of each experiment the heart was excised with the electrodes in place and the relative relationship of the needles confirmed. Each of the 768 channels was sampled with a frequency of 960 samples/sec and filtered to provide a bandwidth of 0.07-170 HZ (7). Local activation time was defined as dominant negative deflection in the unipolar electrogram, an automated algorithm that determines slope of -5V/sec was used to annotate detection in similar electrograms in previous work (7,8) We focused our analysis on the spontaneous PVCs that triggered ventricular arrhythmias.