Combined Bispectral and Bicoherency approach for Catastrophic Arrhythmia Classification

Quantitative classification of cardiac arrhythmia is an important tool in ICU and CCU that enables on line monitoring of the cardiac activities. Among fatal arrhythmias are atrial fibraliation (AF), ventricular tachycardia (VT), and ventricular fibrillation that require special algorithms for detection and so for direct medical actions. In this paper, a combined bispectrum and bicoherency classification algorithm is introduced. It is based on extracting diagnostic features from the bispectrum contours and the bicoherency indices that better describe the arrhythmia. A simple classification scheme utilizing these features showed notable sensitivity and specificity. The obtained results are found comparable to the state of the art algorithms with the ability of being integrated for on line monitoring and classification.