Delineation of QRS-complex, P and T-wave in 12-lead ECG

An accurate and efficient method of detection and delineation of QRS-complexes, P-waves and T-waves in 12-lead ECG is presented in this paper. The delineation process is completely automatic and is based on detections done on the basis of adaptive quantized threshold of the QRS and non-QRS feature signal. The feature extraction is done using a modified definition of slope of the ECG signal. Due to the inclusion of slope and threshold, of ECG and feature signal respectively, the filtering of the raw ECG signal for noise removal and baseline drift removal become pre-requisite. The delineation performance of the proposed algorithm is validated using original ECG recordings from the dataset-3 of the CSE multi-lead measurement library. The delineation results obtained clearly indicate a high degree of agreement with the manual annotations done by the referees of CSE dataset-3.