A four-lead real time arrhythmia analysis algorithm

Abstract

Multi-lead ECG analysis is an effective method to improve detection accuracy and reduce false positive alarms. A four-lead arrhythmia analysis algorithm that processes up to four ECG leads is described. QRS complexes detected in each lead are compared to determine their acceptability. Once the QRS complexes are found acceptable, the information from all the acceptable leads is synthesized to detect and classify them. The arrhythmia analysis section then calculates heart rate and detect the arrhythmia. The AHA and MIT-BIH databases (DB1) as well as four-lead Mindray databases (DB2∼DB5) were used to evaluate the performance of the algorithm. The results for DB1 show QRS detection and classification performance against the standard databases. Testing with data from DB2 and DB3, shows that QRS detection and classification accuracies when using four-lead analysis is superior to the results obtained using fewer leads. For DB2∼DB5 the false and missed lethal alarms were reduced more than 75% and 90% respectively using the four lead algorithm compared to two lead analysis.