Real-time detection of electrocardiograph peaks: A genetic algorithm based approach

Real-time detection of QRS complexes is quintessential for continuous electrocardiograph (ECG) monitoring. In this paper, a real-time ECG peaks detection algorithm is proposed. The paper proposes a new genetic algorithm (GA) based adaptive filter (GA-AF) and an effective real-time peak detection algorithm, which is coupled with search-back mechanism for missed peaks. The proposed methodology is tested on the standard MIT-BIH Arrhythmia database. The GA-AF based QRS detector gives good detection performance by taking an average duration of 0.3s for processing a signal block of 0.83s duration. The proposed methodology achieves sensitivity (%Se) and positive predictivity (%P+) of 98.97% and 99.70% respectively.