A deep learning-powered system for detecting 14 chest diseases from X-ray images using ResNet50, Grad-CAM visualizations, and the NIH Chest X-ray dataset. Designed for interpretability and high diagnostic accuracy, even with class imbalance — using LSE pooling and Binary Cross Entropy loss.
ai computer-vision deep-learning grad-cam artificial-intelligence medical-imaging artificial-neural-networks resnet-50 chest-xray nih-dataset nih-chest-x-ray fastapi ai-healthcare mulit-label-classification
-
Updated
May 27, 2025 - HTML