Analysis of abnormalities in cardiac arrhythmia based on 12 - LEAD electrocardiography

Abstract

Myocardial Infarction otherwise called heart attack occurs in human beings when blood flow decreases or stops to a part of the heart which in turn damages the heart muscle. Prediction of abnormalities in cardio arrhythmia disease is done by using standard 12-lead Electrocardiography (ECG) signals, which also detects Posterior Myocardial Infarction (PMI). The QRS complex is the merged output of different parts of graphical deflection seen on a typical Electro Cardio Gram (Electrocardiography). The main purpose of the paper is to monitor and analyze particularly the R_peak upward deflections from the QRS complex. Denoising the ECG signal is done by butter worth filter. The denoised signals are used to detect R peaks and image plotting is done by segmentation. R peak images are used to classify the abnormalities in Myocardial Infarction (MI) with the help of the CNN image processing technique. The publicly available PTB diagnostic dataset is used to classify the abnormalities in PMI. The detection of the R peaks is used to guide Cardiologists must advance the Percutaneous Coronary Intervention treatment. Prediction has been done using probability weighted average method. Troponin level has been calculated to evaluate a person's health condition which also supports in close prediction of diseases and abnormalities. From experimental analysis of proposed Probability weighted average method in troponin level (PWAMT), the accuracy scores in the ensemble model were found to be 86 % respectively. The running of algorithm took 250 s–300 s to execute the program and display the prediction results.