Meta Heuristic Based Optimized Intelligent Framework for Kidney Disease Detection Using Deep Learning

Chronic Kidney Disease is a slow corrosion in kidney function that can happen in kidney damage or failure. The difficulty of making a diagnosis increases as the illness develops. Early identification and rapid intervention could be facilitated by evaluating different phases of the disease using data from normal medical appointments. In order to better diagnose kidney illness, this research suggests a novel framework that combines Deep Learning with meta-heuristic optimization methods. To improve the efficiency of deep learning models, they used two meta-heuristic optimization methods for Feature Selection, namely Ant Lion Optimization and Bat Inspired Algorithm. The investigation was conducted using two deep learning models, namely LSTM and RNN. Results in terms of accuracy, precision, recall, and F1-score are achieved by combining deep learning models with meta-heuristic optimization strategies. According to the results of the trial, the DL strategy that incorporates FS methods yields superior performance. With a recall of 98.64%, accuracy of 98.83%, and precision of 97.70%, the BLA-RNN model performs better than the ALO-LSTM, ALO-RNN, BLA-LSTM, and BLA-RNN models in every category.