Title: An approach to determine the amplitude features from ECG records

Authors: H.K. Chatterjee; R. Gupta; M. Mitra

Addresses: ECE Department, Camellia School of Engineering and Technology, Nadibhag, P.O. Kagipara, Barasat, Calcutta 700124, West Bengal, India ' Department of Applied Physics, University of Calcutta, 92, Acharya Prafulla Chandra Road, Calcutta 700009, West Bengal, India ' Department of Applied Physics, University of Calcutta, 92, Acharya Prafulla Chandra Road, Calcutta 700009, West Bengal, India

Abstract: In this work, an electrocardiogram feature extraction algorithm is developed by analysing ECG signal in time domain. At first all R peaks are determined using amplitude and slope based criteria. Base line modulation was removed from ECG signal by making the midpoint of successive R to R regions to lie on the same horizontal line and thereby adjusting the values of all intermediate points proportionately. With respect to detected QRS peak positions, other wave peaks are determined based on slope and then amplitude based search in the respective searching zone. P, Q, R, S and T wave peak positions, their onset and offset points and their heights w.r.t. the baseline are calculated. The algorithm is validated with mit-db and ptb-db ECG data files over a number of class of abnormalities, and achieved an average variance and percentage error of 0.0020 and 1.88% for wave peak height computation.

Keywords: ECG signals; QRS detection; amplitude features; physionet; ptb-db; mit-db; baseline; onset points; offset points; electrocardiograms; feature extraction.

DOI: 10.1504/IJBET.2013.056506

International Journal of Biomedical Engineering and Technology, 2013 Vol.12 No.2, pp.130 - 146

Received: 10 Nov 2012
Accepted: 11 Jul 2013
Published online: 27 Sep 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article