Enhancement of the modified p-spectrum for use in real-time QRS complex detection

Abstract

QRS complex (heart beat) detection is frequently used by physicians to detect abnormal electrical activities (such as the cardiac arrhythmia, tachycardia, and bradycardia) in the heart. Algorithms for real-time detection of cardiac arrhythmia are critical, as ambulances and critical care facilities often require heart rate data for in-situ diagnosis of a patient. In this paper, a new algorithm based on the modified p-spectrum transform [1] for QRS complex detection is proposed. The strengths of the original transform are inherited, allowing for high accuracy detection without requiring a priori knowledge of the electrocardiograph (ECG) signal. When tested using the MIT-BIH Arrhythmia Database [2], the new algorithm yields identical performance characteristics and require substantially less computation time. All 48 1/2-hour ECG datasets were tested and the proposed approach outperformed [1] by a factor of up to 3.5 at high sampling frequencies.