Is waveform analysis a viable consideration for implantable devices given its computational demand?

Morphometric analysis of intracardiac waveforms provides a potentially valuable feature for identification of abnormal cardiac activation. Its use as a companion to rate analysis for rhythm classification has been suggested for implantable cardioverter-defibrillators (ICDs). Preliminary studies have shown improved specificity of diagnosis which could be expected to reduce false shocks. Power consumption is directly related to computational burden, thus abbreviated algorithms are required if this technology is to be considered viable for battery-operated devices. In this study, three fast algorithms were tested on 16 patient recordings containing sinus rhythm, ventricular tachycardia, and ventricular fibrillation in the same patient, and compared to correlation waveform analysis (CWA) of the same rhythm passages. Results showed that fast algorithms, with reduced computational load were equally robust and yielded equivalent results to CWA. Promise for such algorithms in future ICDs remains a viable possibility