Explainable Artificial Intelligence (XAI) is crucial in medical diagnosis to enhance transparency, interpretability, and trust in AI-driven decision-making. This study explores the application of XAI techniques for Chronic Kidney Disease (CKD) prediction and lung disease detection. For CKD, a machine learning model was developed using patient clinical data, with SHAP (SHapley Additive exPlanations) employed to identify the most influential features affecting predictions. The results demonstrate that attributes such as serum creatinine, blood urea, and hemoglobin levels significantly impact CKD risk assessment. For lung disease detection, a deep learning model was trained on chest X-ray images, and Grad-CAM (Gradient-weighted Class Activation Mapping) was applied to generate visual explanations, highlighting the critical regions influencing model decisions. Experimental results show that both methods improve the interpretability of AI predictions, aiding healthcare professionals in understanding and validating the model’s reasoning. The study highlights the importance of integrating explainability into medical AI models to ensure reliability, facilitate clinical adoption, and enhance patient trust. Future work includes expanding the dataset and exploring additional XAI techniques to further improve model transparency.
Shah et al. (Fri,) studied this question.