Uso de los Términos Cruzados de la Distribución Choi-Williams para Detectar Actividad Tardía Ventricular

Abstract

Most methods used for malignant ventricular tachycardia risk prediction make use of the late ventricular potentials as an anomalous factor in the averaged electrocardiographic signal analysis of post-infarct patients. However, the detection of the late potentials is a very complicated task, due to the fact that their amplitude and frequency are very close to those of the noise. Nevertheless, in this work, an indirect method of late potential detection is proposed, using the values od the cross-terms that ere produced in the time-frequency plane generated by the Choi-Williams distribution function. This indirect method eliminates the difficulties associated with the traditional methods, such as the classical temporal domain method. This analysis was used on 132 averaged electrographic signals, of which 59 correspond to post-infarct patients and 73 correspond to healthy individuals. The predictive results of the proposed method show a substantial improvement over the classical temporal domain method results.