AI Powered Decision Support Systems for Healthcare Enhancing Diagnosis and Treatment with Deep Learning

This research presents a novel hybrid decision support system with artificial intelligence based on Swin Transformer and Graph Neural Networks (GNNs) to enhance diagnostic accuracy and personalized treatment in medical care. Swin Transformer efficiently extracts hierarchical features from medical imaging data, while GNNs specify patient-specific interrelations from structured Electronic Health Records (EHRs). Combining spatial and relational knowledge, the system provides better predictive performance. Implemented using Python with deep learning libraries PyTorch, the hybrid model is evaluated using real-world healthcare datasets. Experimental results show that the proposed method shows an accuracy rate of 92%, a 7–10% increase compared to individual models like CNNs, GNNs, and Swin Transformers. The performance improvement leads to more accurate disease detection and better patient outcomes. Additionally, the model provides good scalability and usability in real-world applications, especially health camp resource optimization. These promising results make the system a strong and interpretable AI platform for clinical deployment.