๐‘๐ž๐ฌ๐ž๐š๐ซ๐œ๐กไธจLLM-based Radiology Report Generation by Liu Chang๐‘๐ž๐ฌ๐ž๐š๐ซ๐œ๐กไธจLLM-based Radiology Report Generation by Liu Chang

๐‘๐ž๐ฌ๐ž๐š๐ซ๐œ๐กไธจLLM-based Radiology Report Generation

Liu Chang

Liu Chang

AI Researcher & First Author
๐Œ๐จ๐ญ๐ข๐ฏ๐š๐ญ๐ข๐จ๐ง & ๐‚๐ก๐š๐ฅ๐ฅ๐ž๐ง๐ ๐ž: General-purpose LLMs lack the domain alignment needed for radiology report generation. We combined contrastive in-domain learning with coarse-to-fine report refinement guided by X-ray features.
๐‘๐ž๐ฌ๐ฎ๐ฅ๐ญ๐ฌ & ๐•๐ž๐ง๐ฎ๐ž๐ฌ: Improved report generation on IU X-Ray and MIMIC-CXR. Published at AAAI 2024.
๐‹๐ข๐ง๐ค๏ผšhttps://ojs.aaai.org/index.php/AAAI/article/download/29826/31434
Like this project

Posted Sep 15, 2026

Combined in-domain learning with coarse-to-fine LLM report refinement on IU X-Ray and MIMIC-CXR. Published at AAAI 2024.