7 papers
Towards end-to-end LLM-based censoring-aware survival analysis
Yishu Wei, Hexin Dong, Yi Lin +3
Objective: Survival analysis is central to medical prediction, yet large language models (LLMs) are rarely used as end-to-end survival models because censoring prevents straightfor…
Reinforcement Learning Improves LLM Accuracy and Reasoning in Disease Classification from Radiology Reports
Yishu Wei, Yi Lin, Adam Flanders +2
Accurate disease classification from radiology reports is essential for many applications. While supervised fine-tuning (SFT) of lightweight LLMs improves accuracy, it can degrade…
Curation and Extraction of Drug-Related Entities from Reddit Platform
Zewei Wang, Zihan Xu, Yishu Wei +2
Physicians learn primarily about illicit drugs from clinical overdose cases, limiting their understanding of real-world usage. Meanwhile, drug users share first-hand experiences on…
RSNA Large Language Model Benchmark Dataset for Chest Radiographs of Cardiothoracic Disease: Radiologist Evaluation and Validation Enhanced by AI Labels (REVEAL-CXR)
Yishu Wei, Adam E. Flanders, Errol Colak +35
Multimodal large language models have demonstrated comparable performance to that of radiology trainees on multiple-choice board-style exams. However, to develop clinically useful…
A Multi-Stage Large Language Model Framework for Extracting Suicide-Related Social Determinants of Health
Song Wang, Yishu Wei, Haotian Ma +10
Background: Understanding social determinants of health (SDoH) factors contributing to suicide incidents is crucial for early intervention and prevention. However, data-driven appr…
CXR-LT 2024: A MICCAI challenge on long-tailed, multi-label, and zero-shot disease classification from chest X-ray
Mingquan Lin, Gregory Holste, Song Wang +30
The CXR-LT series is a community-driven initiative designed to enhance lung disease classification using chest X-rays (CXR). It tackles challenges in open long-tailed lung disease…