Covering rough set-based classification for cardiac arrhythmia

Abstract

The objective of this work is to use data processing methods to unravel the biomedical difficulties of detecting a selection of arrhythmia conditions from patient's electrocardiograph (ECG) signals. ECG graphical signal record waveforms square portion analysed supported morphological variations between irregular waves and regular waves. The Pan-Tompkins (PT) method is applied for takeout P, R, Q, S and T morphological peaks and wavelet transform (WT) temporal features are tested on five categories of graphical record ECG signals. During this paper, we tend to propose the covering rough set (CRS)-based classification method for classification of heartbeats to sign interior cardiac arrhythmia in ECG signals. The proposed classification system is tested using ECG records in Physiobank databases and also the results were compared to those from many prior studies. Experimental results show that our proposed approach in truth diagnoses heart cardiac arrhythmia. The experimental results show that the proposed system outperforms other classifiers.