A collection of resources on applications of multi-modal learning in medical imaging.

Foundation models based medical image analysis

Awesome radiology report generation and image captioning papers.

Official implementation of "UniMedVL: Unifying Medical Multimodal Understanding and Generation through Observation-Knowledge-Analysis" - A unified medical vision-language model that integrates multimodal understanding and generation capabilities.

[ACL 2023] ORGAN: Observation-Guided Radiology Report Generation via Tree Reasoning

[ACL 2025] RADAR: Enhancing Radiology Report Generation with Supplementary Knowledge Injection

[EMNLP 2023 Findings] RECAP: Towards Precise Radiology Report Generation via Dynamic Disease Progression Reasoning

Code used for the MLMI 2021 paper Clinically Correct Report Generation from Chest X-Rays Using Templates

[EMNLP 2024 Findings] ICON: Improving Inter-Report Consistency in Radiology Report Generation via Lesion-aware Mixup Augmentation

Medical Report Generation And VQA (Adapting XrayGPT to Any Modality)

[IJBHI 2024] This is the official implementation of CAMANet: Class Activation Map Guided Attention Network for Radiology Report Generation accepted to IEEE Journal of Biomedical and Health Informatics (J-BHI), 2023.

This is the official implementation of MvKeTR: Chest CT Report Generation with Multi-View Perception and Knowledge Enhancement accepted to IEEE Journal of Biomedical and Health Informatics (J-BHI), 2025.

AI-powered Chest X-ray report generation app using VLM (Swin-T5) and LLM (LLaMA-3) for multilingual Q&A and medical education support.

GPT-2 based medical reports generator for X-ray images in Czech.

Medivance.AI is a cutting-edge, all-in-one AI healthcare platform that transforms the way patients, doctors, and healthcare organizations interact with medical data.

Resources on the use of multimodal learning in medical imaging.

The primary objective of this work is to develop an innovative system capable of providing explainable brain tumor detection.

Source code for our AMIA 2025 paper (oral): "A Clinically-Informed Framework for Evaluating Vision-Language Models in Radiology Report Generation: Taxonomy of Errors and Risk-Aware Metric"

Bio-J.A.R.V.I.S. is a tool for generation of medical report texts using generative AI

A Flutter-based mobile application designed to digitize and streamline pre-hospital patient care reporting. It replaces cumbersome paper forms, reducing administrative overhead and improving the accuracy and speed of emergency medical data collection.