Resources

We introduce MIMIC-Ext-MIMIC-CXR-VQA, a complex, diverse, and large-scale dataset designed for Visual Question Answering (VQA) tasks within the medical domain, focusing primarily on chest radiographs.

A benchmark dataset offering hourly records over a 90-day period for 50,920 ICU subjects, including dynamic pulmonary function data and a spectrum of covariates for respiratory intervention analyses.

MS-CXR is a new dataset containing 1162 Chest X-ray bounding box labels paired with radiology text descriptions, annotated and verified by two board-certified radiologists.

We present EHRXQA, the first multi-modal EHR QA dataset combining structured patient records with aligned chest X-ray images. EHRXQA contains a comprehensive set of QA pairs covering image-related, table-related, and image+table-related questions.

We propose ReXPref-Prior, an adapted version of MIMIC-CXR where GPT-4 has removed references to prior exams from both findings and impression sections of chest X-ray reports.

RadGraph2 is a dataset of 800 chest radiology reports annotated using a fine-grained entity-relationship schema, which captures key findings as well as mentions of changes that occurred in comparison with the previous radiology studies.

We present EHRXQA, the first multi-modal EHR QA dataset combining structured patient records with aligned chest X-ray images. EHRXQA contains a comprehensive set of QA pairs covering image-related, table-related, and image+table-related questions.

Me-LLaMA is a family of large language models for medical applications trained using clinical text with LLaMA2 models as the base. We release model weights for the foundation models as well as the chat-enhanced models.

Asclepius: Publicly Available Clinical Large Language Models with Synthetic Clinical Notes Asclepius-R: A instruction-finetuned large language model with MIMIC-III clinical notes

Fine tune transformer-based neural networks to deidentify clinical text data.