Computer based two-dimensional spectral estimation for the detection of prearrhythmic states after hypothermic myocardial preservation

A computer based two-dimensional spectral estimation technique has been developed to detect prearrhythmic states of the heart after hypothermic myocardial preservation with or without potassium cardioplegia. The new algorithm is based on unconstrained minimization of the estimated covariance recursion error. The parameters for spectral analysis are estimated directly from intracardiac data in a manner which combines high resolution with robustness in the presence of nonstationarities. It is shown that by using this technique, the projections of the velocity vector of cardiac propagating wavefronts onto a multitip electrode array may be determined, even though electric intramyocardial data are very short and nonstationary. The prearrhythmic state may be observed at a time when limb and epicardial ECG waveforms are normal. This is of potential significance since certain types of prearrhythmic patterns appear more likely to degenerate into ventricular arrhythmias or ventricular fibrillation.