Title: An Interpretable Machine Learning and Deep Learning Framework for Early Prediction of Chronic Kidney Disease Using Clinical Data
Authors: Salah Aldin Shadi Aldaya, Duha Khalil Ahmed Aldaya, Samy S. Abu-Naser
Volume: 9
Issue: 12
Pages: 97-105
Publication Date: 2025/12/28
Abstract:
Chronic Kidney Disease (CKD) is a progressive and often asymptomatic condition that poses a significant global public health burden due to its association with high morbidity and mortality rates. Early detection of CKD is crucial for preventing disease progression and reducing the risk of severe complications. In recent years, machine learning (ML) and deep learning (DL) techniques have demonstrated promising capabilities in medical decision support systems; however, their limited interpretability remains a major barrier to clinical adoption. In this study, an interpretable machine learning and deep learning framework is proposed for the early prediction of chronic kidney disease using clinical data. The framework is developed and evaluated using a publicly available CKD dataset comprising 400 patient records with 24 clinical and laboratory features. Comprehensive data preprocessing steps, including missing value imputation, categorical feature encoding, and feature normalization, are applied to enhance data quality and model performance. Multiple machine learning models, including Logistic Regression, Support Vector Machine, Random Forest, and Gradient Boosting, are implemented and compared with a fully connected deep learning model. Model performance is assessed using clinically relevant evaluation metrics, including accuracy, precision, recall, F1-score, and the area under the receiver operating characteristic curve (ROC-AUC). To ensure transparency and clinical interpretability, explainable artificial intelligence (XAI) techniques such as SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME) are employed to identify the most influential clinical features contributing to CKD prediction at both global and individual patient levels. Experimental results demonstrate that the proposed framework achieves high predictive performance while maintaining interpretability, highlighting the clinical relevance of key risk factors such as serum creatinine, blood urea, hemoglobin levels, and hypertension status. The findings suggest that the integration of interpretable ML and DL models can provide reliable and transparent decision support for early CKD diagnosis, potentially assisting clinicians in timely intervention and improved patient outcomes.