Two-Phase Heart Disease Diagnosis System Using Deep Learning
Abstract
Heart disease diagnose mechanism can differ, depending on the type of heart disease. However, the common routine which most medical practitioners use to make a diagnosis consisting of physical examination using clinical data followed by a specific medical test like Electrocardiogram, Echocardiogram, Cardiac CT Scan, Cardiac MRI Study, Stress Testing, Cardiac Catheterization, and Electrophysiology Study. Out of these specific tests, Electrocardiogram is a widely used test that can detect information about the heart rhythm and significant indications about structural heart disease. This work presents a scheme that first uses clinical data to identify the chances of cardiovascular problems. In case of positive chances, the ECG signals are used to classify heart rhythm for diagnosing the particular type of heart disease. In this integrated approach, deep neural network-based methods are utilized and results confirm the robustness of the proposed model.
Keywords: Cardiovascular disease, ECG, Arrhythmia, Deep Learning, Convolutional Neural Network