An Intelligent REMORA Optimized RNN With LSTM Mechanism for Detection of Atherosclerosis

Abstract

Atherosclerosis is a medical condition where plaque accumulates in the arteries. It is important to detect and intervene early to minimize the negative effects of atherosclerosis on a person's health. The World Health Organization (WHO) states that 17.9 million deaths globally are caused by cardiovascular diseases (CVDs), the primary cause of death in the country. Several medical innovations and advancements were being explored in the field of atherosclerosis. The proposed research aims to build a system based on artificial intelligence (AI) automatic framework for an accurate detection and classification of Atherosclerosis using a hybrid REMORA Optimized Recurrent Neural Network with Long Short. Term Memory (RO-RNN-LSTM) method. In this work, various techniques for feature selection, preprocessing, and classification are implemented to achieve more accuracy. The min-max normalization followed by, the relevant features from the preprocessed data are extracted using the innovative REMORA Optimization method. In order to maximize accuracy and minimize classifier time, the dimensionality of features was reduced in this work. The publicly available datasets such as Cleveland, Kaggle (CVD) and Stulong are used to evaluate the proposed algorithms. The prediction findings indicate that utilizing the RO-RNN-LSTM algorithms improves accuracy up to 98.5%, sensitivity to 98.7%, specificity to 95.4%, recall to 98.7% and f1-score to 98.7%. The RO-RNN-LSTM exhibits superior performance outcomes while processing large dimensional medical datasets, according to the overall comparison investigation and analysis.