Abstract:

The global prevalence of chronic kidney disease (CKD) is increasing at an alarming rate. Many cases remain asymptomatic, and guideline-based monitoring for early detection is often underutilized. Computer-Aided Diagnosis (CAD) systems, particularly those leveraging deep learning, offer promising solutions for the early prediction of CKD due to their superior classification capabilities. This study explores the use of various clinical features associated with CKD and implements seven advanced deep learning algorithms—Artificial Neural Networks (ANN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Bidirectional LSTM, Bidirectional GRU, Multilayer Perceptron (MLP), and Simple Recurrent Neural Network (RNN)—for effective prediction and classification of CKD. These models, grounded in artificial intelligence, were trained using five distinct feature extraction approaches on preprocessed CKD datasets. Performance was evaluated using metrics such as accuracy, precision, recall, loss, and validation loss. Additionally, the study assessed computation time, prediction ratios, and Area Under the Curve (AUC), alongside statistical significance tests to comprehensively compare model performance.

Description:

.