PREDICTING PATTERNS OF EPICARDIAL POTENTIALS DURING VENTRICULAR-FIBRILLATION

Ventricular fibrillation (VF) is a fatal cardiac arrhythmia, character ized by uncoordinated propagation of activation wavefronts in the vent ricular myocardium. Short-term predictions of epicardial potential fie lds during VF in pigs were attempted using linear techniques, and pred iction accuracy was measured at various stages during sustained episod es, VP was induced in five pigs via premature electrical stimulation. Unipolar electrograms were recorded from an epicardial array of 506 el ectrodes in a 22 x 23 array with 1-mm spacing, Optimal spatial basis f unctions (modes) and time-varying weighting coefficients were found us ing the Karhunen-Loeve decomposition, Linear autoregressive (AR) model s incorporating the dynamics of only a few spatial modes led to predic ted patterns that were qualititatively similar to observed patterns. P redictions were made 0.256 s into the future, based on 0.768 s of past data, over an area: of approximately 5 cm(2) on the ventricular epica rdium. The mean squared error of predictions varied from as much as 1. 23 to as little as 0.14, normalized to the variance of the actual data , Inconsistency in long-term forecasts is partly due to the limitation s of linear AR models, Changes in predictability, however, were consis tent. Predictability varied inversely with spatial complexity, as meas ured by the mean squared error of a five-mode approximation. Predictab ility also increased significantly during the first minute of VF.