Journal of Theoretical and Applied Information Technology 31 st December 2022. Vol.100. No 24 © 2022 Little Lion Scientific ISSN: 1992-8645 www.jatit.org E-ISSN: 1817-3195 4771 CORONARY ARTERY DISEASE PREDICTION BASED ON OPTIMAL FEATURE SELECTION USING IMPROVED ARTIFICIAL NEURAL NETWORK WITH META-HEURISTIC ALGORITHM D.VETRITHANGAM 1 , V. SENTHILKUMAR 2 , NEHA 3 , A. RAMESH KUMAR 4 , P.NARESH KUMAR 5 , MRADULA SHARMA 6 1 Associate Professor , Department of Computer Science & Engineering , Chandigarh University , Punjab, India. 2 Associate Professor, Department of Mechanical Engineering, SRM TRP Engineering College, Tamilnadu,India. 3 Assistant Professor , Department of Computer Science & Engineering ,Chandigarh University , Punjab, India. 4 Professor, Department of Mechatronics Engineering, K.S.Rangasamy College of Technology, Tamil Nadu,India. 5 Assistant Professor, Department of Computer Science and Engineering, KG Reddy College of Engineering and Technology, Telangana,India. 6 Assistant Professor, Department of Computer Science and Information Technology, Japyee Institute of Information Technology,Uttarpradesh,India. E-mail: 1 vetrigold@gmail.com, 2 trpvsk12@gmail.com, 3 neha.arya35@gmail.com, 4 arameshkumaar@gmail.com, 5 pnrshkumar@gmail.com, 6 mradulasharma86@gmail.com ABSTRACT Scientific breakthroughs in understanding the etiology of coronary artery disease (CAD) will allow for more accurate coronary artery disease (CAD) diagnosis and treatment techniques. Coronary Artery Disease (CAD) is a type of cardiovascular disease in which atherosclerotic plaques in the coronary arteries cause myocardial infarction or sudden cardiac death. In medicine, disease prediction based on Artificial Neural Networks (ANN) plays a significant role in enhancing the reliability of general population health care. So, our main goal is to propose an improved artificial neural network model in conjunction with a Meta- heuristic algorithm that works with distinct types of CAD datasets with good accuracy. The system selects the most relevant or similar features from the raw dataset; this feature selection is achieved by the Meta- heuristic algorithm. This model uses 18 input nodes, 18 hidden nodes, and 1 output node in an 18-18-1 multilayered feed-forward network architecture, which is the best network for the prediction of CAD with the selected dataset. When using different methodologies on datasets dealing with coronary artery disease (CAD), the results may vary. Efficient medical diagnosis and analysis are important in selecting the important features. The Cleveland Heart Disease dataset, obtained from the UCI repository, was used in this paper; it contains 37079 person data records with 50 attributes. The proposed Improved Artificial Neural Network model with Meta-heuristic Algorithm results in 97.63% Sensitivity, 97.5% Accuracy, and 97.35 % Specificity. Keywords: Coronary Artery Disease, Deep Learning, Risk Factor Meta-Heuristic Algorithm And Artificial Neural Networks. . 1. INTRODUCTION Heart disease (HD) could be an essential term for different types of diseases, factors and abnormalities that affecting the heart and also the blood vessels. The symptoms rely upon the particular variety of these diseases such as heart failure, hypertensive heart disease, coronary artery diseases, stroke, cardiomyopathy and congenital disease, heart arrhythmia etc. According to the World Health Organization, cardiovascular disease (CVDs) lead to thirty one percentages of all global deaths, with the number of deaths from CVDs