Resources

This repository releases the EchoGraph-annotated ECHO-NOTE2NUM dataset, containing 45,794 echocardiography reports with comprehensive entity and relation annotations. Each report from the ECHO-NOTE2NUM dataset has been automatically annotated using …

This dataset, Swiss-Mammo, contains 28 manually constructed German mammography reports, each paired with an English translation. The reports are stratified across BI-RADS categories 0 through 6, with four reports per category. All reports were manua…

Recent progress in Large Vision-Language Models (LVLMs) has enabled promising applications in medical tasks such as report generation and visual question answering. However, existing benchmarks focus mainly on the final diagnostic answer, offering l…

We present MIMICEchoQA, a benchmark dataset for echocardiogram-based question answering, built from the publicly available MIMIC-IV-ECHO database. Each echocardiographic study was paired with the closest discharge summary within a 7-day window, and …

Radiology reports are essential for clinical care but pose challenges for automated processing due to their unstructured nature. Existing datasets like RadGraph-1.0 focus narrowly on chest X-rays (CXR), limiting their applicability. We introduce Rad…

The ER-Reason dataset is a benchmark designed to evaluate LLM-based clinical reasoning and decision-making in the emergency room (ER), a high-stakes setting where clinicians make rapid, consequential decisions across diverse patient presentations an…

This dataset contains physiological signals and self-reported measures collected during a controlled stress protocol. Sixty-six healthy adults participated in standardized experimental sessions consisting of a relaxation phase followed by stress ind…

We present RadVLM, a compact (7B) multitask conversational foundation model designed for CXR interpretation. Its development relies on the curation of a large-scale instruction dataset comprising over 1 million image-instruction pairs containing bot…

This project provides MIMIC-Patient, a structured, per-patient representation of the MIMIC-III database designed for large language model (LLM)–based clinical question answering. For each of 500 admissions, we reconstruct the patient’s episode into …

This original dataset contains physiological signals collected during structured acute stress induction and aerobic and anaerobic exercise sessions using a wearable device. Blood volume pulse, motion-based activity, skin temperature, and electroderm…