Artificial intelligence in improving disease diagnosis

Abstract

Every year, cardiovascular disease (CVD) accounts for one-third of the world's fatalities. Premature death affects 43% of the poor, compared to 7% of the rich. The cause is lifestyle-related disorders such as diabetes and obesity. There was a low rate of premature mortality, which highlights the need for early cardiac illness identification. It is crucial to combine biochemical and clinical data for the early detection of cardiac disease. By giving cardiologists access to state-of-the-art data analysis and therapeutic decision-making tools, artificial intelligence (AI) treatments can potentially revolutionize the field. AI tools, such as deep learning (DL) and machine learning (ML), can aid medical professionals in learning more complicated and massive data sets. Early detection of heart disease is very necessary. In this paper, we have developed two-stage experiments on an integrated five-heart disease dataset from the UCI-ML warehouse. First, we tested 13 ML and DL algorithms on the integrated dataset. In the second stage, we applied these conventional classifiers to the featured dataset obtained by applying sequential feature selection and observed the enhanced predictive outcomes with the same integrated dataset as 97.06% accuracy, 98.81% precision, 97.08% sensitivity, 97.01% specificity, and 97.94% F1 score that justifies the novelty of the proposed model. Finally, ROC curves with AUC values were added to the study's empirical analysis, and an improved AUC of 0.987 was recorded. This proposed work might help physicians diagnose heart diseases.

Topics & Concepts