Feature Extraction and Classification of ECG Signal Based on The Standard Extended Wavelet Transform Technique: Cardiology Based Telemedicine

Abstract

For early detection of cardiac abnormalities, an ECGs cardiac signal is relied upon due to it includes a lot of information that can be utilized for heart disease classification. The ECG signal is too sensitive to various types of noise since it is low frequency and has a small amplitude, these noises decrease the diagnostic accuracy and may outcome in a wrong decision by the clinician. Therefore, rejecting the ECG signal is a necessary condition for successful diagnosis of heart attacks. In this manuscript, the standard extended of Discrete Wavelet Transformc called Dual-Tree Complex Wavelet Transform (Dual-Tree (CWT)) method is utilized to denoise the noisy ECG signal and extract the key features followed by the implementation of the peak detection algorithm. The quality is measured based on performance metrics, and an increase in Signal to Noise Ratio ( SNR) is achieved utilizing the technique. The heart rate (HR) calculation is in line with the gold standard of the various benchmark databases utilized for the proposed procedure and precise heart failure has been calculated.