Classification of Coronary Artery Diseases using Electrocardiogram Signals
Muhammad Umar Khan, Sumair Aziz, Syed Zohaib Hassan Naqvi, Abdul Rehman
Abstract
Coronary Artery Diseases (CAD) are the leading cause of adult mortality and morbidity globally. Eelectrocardiogram (ECG) is the most promising biophysical signature for the study of cardiac diseases. In this study, we proposed a signal processing approach to predict Coronary artery disease using raw ECG signals of 9-12 minutes. The raw ECG recording is first pre-processed and segmented using Empirical Mode Decomposition (EMD) by selecting intrinsic mode function (IMF) 2-5. The features that best classify the data are Skewness, Kurtosis, Shape-factor, Impulse Factor, Marginal Factor, Energy, Root sum square, Spectral Entropy, Energy Entropy, Quantile, and Higuchi Fractal Dimension. The preprocessed signal is then fed to the Support Vector Machine classifier. The system achieves 95.5% accuracy on Self-Collected data. The proposed system will help Cardiologists to make effective decisions about the treatment.