Deep Learning Based Detection of Cardiovascular Defect Patients

Abstract

Globally, cardiovascular diseases (CVDs) have surpassed all others as a major killer in the last few years. The symptoms of CVDs are often mild at first, but they might worsen with time. Upon beginning CVD, most people encounter a number of symptoms, including fatigue, difficulty breathing, edema in the ankles, and fluid retention. The majority of cases of CVD include arrhythmia, cardiomyopathy, mitral regurgitation, angina, and congenital heart defects (CHDs). The use of cardiac magnetic resonance imaging (CMR) for diagnosis, disease monitoring, treatment planning, and CVD prediction is on the rise among diagnostic modalities. Despite the many benefits of CMR data, doctors still have difficulties diagnosing CVDs owing to factors like poor contrast, many data slices, etc. A lot of research is being done right now to solve these problems by using enhanced deep learning (DL) techniques for the CVDs diagnosis utilizing CMR data. The research employed the Convolution Neural Network (CNN) algorithm in conjunction with MultiLayer Perceptron to further demonstrate the algorithm's efficacy. It is compared against many alternative DL techniques.