LaurentVeyssier/Chest-X-Ray-Medical-Diagnosis-with-Deep-Learning ? reverse-engineered prompt

Build me a clean notebook based project that can look at a chest X ray and predict the 14 possible pathologies from the dataset, then show a GradCAM heatmap so I can see what part of the image influenced the result.

Use the pretrained DenseNet based model and the provided weights if they are available, and make it easy to run on the sample data already in the repo. I want it to load an image, output the probability for each condition, and clearly highlight the most likely findings in a human friendly way. If there is an evaluation section, include ROC style results and a simple way to check performance on the validation or test set.

Please keep it practical and understandable, with enough comments so someone can follow along, and look up current docs online if you need to make the notebook run smoothly.