R Wave Detection for Real-time Processing using Nonlinear Transform and Dynamic Thresholding

Abstract

Heart rate variability (HRV) is considered important for understanding the activity of autonomic nervous system. In that sense, it is very important to detect R waves in ECG and measure the intervals between consecutive R waves. A method for detecting R wave by thresholding is generally useful but the static thresholding method often fails to detect the R waves when an ECG has very noisy waveform like fetal ECG. In this study, we have proposed a method for pulse detection using nonlinear transform and dynamic thresholding and evaluated our proposed method by computer simulation. In the results, the method with Teager-Kaiser energy operator detects all pulses from a noisy simulation signal from which the method with square operator fails to detect some pulses.