Tumor Location-weighted MRI-Report Contrastive Learning: A Framework for Improving the Explainability of Pediatric Brain Tumor Diagnosis

This repository contains the code for our paper titled: "Tumor Location-weighted MRI-Report Contrastive Learning: A Framework for Improving the Explainability of Pediatric Brain Tumor Diagnosis".

The breakdown of the repository is as follows:

• data: dataset classes and preprocessing functions
• losses: losses and metrics
• model: model-related classes and elements
• explainability.py: the code for extracting model attention maps and computing explainability metrics
• fused_main.py: the main training code for the downstream classification task
• fused_main_global.py: the main training code for the CL framework