A new shock advice algorithm designed to classify ECG signals during cardiopulmonary resuscitation

A shock advice algorithm (SAA) that reliably diagnoses the rhythm during cardiopulmonary resuscitation (CPR) would avoid unnecessary CPR interruptions and increase the probability of a successful resuscitation. Current approaches based on filtering the CPR artifact from the ECG or analyzing the corrupted ECG do not meet the American Heart Association's (AHA) requirements for SAA. This study presents a preliminary design of a SAA to classify the rhythm during CPR. It is based on the analysis of five non-overlapping 3 s segments of the corrupt and the filtered ECG. A total of 290 non-shockable and 89 shockable records were analyzed and a specificity above the 95% AHA goal was obtained for a 89.5% sensitivity, only half a point below the AHA goal.