Chronic Kidney Disease (CKD) Prediction Using Stochastic Deep Radial Basis for Feature Extractiona Residual Neural Network

Chronic Kidney Disease (CKD) is an increasingly common health issue, affecting a significant portion of the global population. Accurate prediction of CKD progression is crucial for early intervention and personalized healthcare. However, existing models often fail to adequately capture the temporal dependencies in disease progression, resulting in suboptimal performance. Current predictive models for CKD are limited by their inability to manage categorical variables effectively and address overfitting while capturing the necessary temporal subtleties. This study aims to address these limitations by developing a novel approach that enhances the accuracy and reliability of CKD progression predictions. The proposed method integrates a unique Automated Stochastic Deep Radial Basis (SDRB) for feature extraction, a Residual Neural Network (ResNet) for classification, and a Bidirectional Long Short-Term Memory (biLSTM) network for temporal prediction. This combination is designed to handle the complexities of CKD data, manage categorical variables, and mitigate overfitting while capturing temporal dependencies.Our model significantly outperforms current state-of-the-art models, achieving 92% accuracy, 91% recall, and 93% precision. The SDRB component improves feature extraction, ResNet enhances classification, and biLSTM effectively captures temporal dependencies, resulting in a robust and reliable CKD progression prediction system.