ViT-ResNet Fusion: An Explainable Hybrid Framework for High-Accuracy Multiclass Lung Disease Classification in Chest X-Rays

The main objective of this study is to develop a hybrid deep learning framework, ViT-ResNet Fusion, for accurate multiclass classification of lung diseases from chest X-ray images, including pneumonia, tuberculosis, COVID-19, and other pulmonary disorders. The framework aims to integrate ResNet for effective local feature extraction with Vision Transformer (ViT) for global contextual modeling, thereby leveraging the complementary strengths of both architectures. Within this framework, specialized models—PulmoFusion X, BronchoViT R, and AeroNet S—are designed to capture comprehensive pulmonary features, bronchial patterns, and subtle aeration abnormalities. Additionally, the study seeks to incorporate explainable AI techniques such as Grad-CAM to enhance interpretability, improve clinical trust, and support radiologists in timely and reliable diagnosis.