Threshold-based Detection of P and T-wave in ECG using New Feature Signal

The presented algorithm detects and delineates both P and T-waves simultaneously and employs a modified definition of slope, ofECG signal, as the feature for detection of ECG wave components.

Abstract

Summary Detection of P and T waves is an important part in the analysis and interpretation of ECG. The presented algorithm detects and delineates both P and T-waves simultaneously. It employs a modified definition of slope, of ECG signal, as the feature for detection of ECG wave components. A number of transformations of the filtered and baseline drift corrected ECG signal are used for extraction of this new modified slope-feature. Five feature-components are combined to derive the final feature signal. Amplitude threshold of the final feature signal is employed for distinguishing P and T waves with respect to already detected QRS-complexes. P-wave detection rate of 96.95% with false positive and false negative percentage of 2.62% and 3.01% has been reported. Similarly, T-wave detection rate of 98.01% with false positive and false negative percentage of 3.08% and 1.93% has been reported.