Improved detection and classification of arrhythmias in noise-corrupted electrocardiograms using contextual information

A novel approach for employing contextual information to process noisy electrocardiograms (ECGs) is described. The method is embedded in HOBBES, an expert arrhythmia analysis system. HOBBES is a post-processor to ARISTOTLE, a classical arrhythmia detector. ARISTOTLE and HOBBES were evaluated using a database of 35 half-hour records containing a mixture of supraventricular and ventricular ectopic activity. The HOBBES-assisted system performed dramatically better than ARISTOTLE alone in processing noisy ECGs.<>