Explainable Deep Learning for Lung Disease Detection on Chest X-Ray Images Using Shapley Additive Explanations (SHAP)

X-ray imaging is an effective tool for detecting lung diseases such as COVID-19 and pneumonia, and its capabilities can be enhanced using artificial intelligence-based systems. This study develops a chest X-ray image classification system using a Convolutional Neural Network (CNN) with the VGG-16 architecture and integrates Shapley Additive Explanations (SHAP) to make the model’s predictions explainable. The model is trained on a dataset of 3,030 X-ray images divided into three classes (COVID-19, pneumonia, and normal), using several configurations; the best performance is achieved with an 80%:20% train-validation split, a learning rate of 0.001, a batch size of 32, and 50 epochs. The trained model attains 95.75% accuracy on the training data and 96.00% on the validation data. SHAP is then applied to generate heatmap visualizations that highlight the image regions most influential to each prediction, thereby improving interpretability and supporting clinical decision-making. The results demonstrate that combining deep learning with SHAP can produce both highly accurate and visually interpretable lung disease detection from chest X-ray images.