Efficient Prediction Model for Cardiovascular Disease Using Deep Learning Techniques

Abstract

Cardiovascular illness is becoming more commonplace every day, which makes early identification of the condition worrisome and essential. Making a cardiac diagnosis is a challenging procedure that has to be finished fast and expertly. The identification and prediction of cardiovascular illness are vital medical responsibilities that help cardiologists appropriately diagnose and treat their patients. Deep Learning (DL) algorithms have found increasing usage in the medical industry because of their ability to recognize patterns in data. By applying machine learning to categorize the occurrence of cardiovascular illness, clinicians can reduce the rate of misdiagnosis. The goal of this work is to develop an ML model for cardiovascular disease (CVD) forecast based on correlated problems. This research develops a model that can accurately anticipate CVD illnesses, which will reduce the fatality rate from these ailments. This work employs a variety DL approaches to compare the results and analysis of the UCI Machine Learning Heart Disease dataset. We used a benchmark dataset of UCI Heart disease prediction for this work, which consists of 14 different heart disease-related parameters. As DL models, there are Convolutional Nueral Network (CNN), Artificial Neural Network (ANN) and Long Short-Term Memory (LSTM). When compared to DL techniques, the result shows that LSTM offers superior prediction accuracy in less time. 91% accuracy was attained using this LSTM technique. This model performs better on training and testing data. The models were fitted to the test dataset as well as trained on the training dataset to see which fared best. The matrices acquired during this process were accuracy, specificity, sensitivity, Area Under Curve (AUC) and the Receiver Operating Characteristic Curve (ROC).