AI for Medical Imaging and Diagnostics

Oury core methodology involves leveraging Transfer Learning from modified state-of-the-art models (including ViT, ResNet, Xception, and Swin Transformer). We use advanced feature engineering, employing multi-stage feature fusion (e.g., Convolutional Sparse Image Decomposition (CSID)) and hybrid optimal feature selection to achieve superior diagnostic performance, consistently reaching high accuracy. We also specialize in Explainable AI (XAI), integrating dedicated attention modules and visualization techniques (e.g., Grad-CAM) to enhance model transparency and clinical trustworthiness. This expertise also extends to complex Multi-Criteria Decision-Making frameworks for disease assessment.