发表机构
Georgia Institute of Technology; University of Florida（佐治亚理工学院; 佛罗里达大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
DS@GT ARC团队针对2026 MEDIQA-CORE-Task-1，提出带任务特定门控的三模态模型融合方法，结合多类嵌入与报告，获全多模态下0.801宏F1，排名第二。
AI 中文摘要
脑肿瘤诊断是一个对时间敏感的过程，患者可能需要等待数周才能获得最终病理报告，这一问题推动了可从多模态输入中对肿瘤亚型进行分类的自动化系统。本文详细介绍了DS@GT ARC团队在2026 ImageCLEFmed MEDIQA-CORE任务1（脑肿瘤亚型分类）中的工作，该任务评估三项胶质瘤分类问题：Level-1分子类型、低级别胶质瘤（LGG）vs高级别胶质瘤（HGG）、WHO分级。我们将预提取的MRI（NeuroVFM）嵌入、组织病理学（Prov-GigaPath）嵌入与自由文本放射学报告相结合，团队探索了两种三模态融合架构、两种报告编码器（RadBERT和Llama-3.1-8B-Instruct），以及一个受生物学启发的后处理阶段。在全多模态条件下，我们达到0.801的平均宏F1值，超过组织者基线的0.796，在代码通过验证的团队中排名第二；跨模态丢弃条件下的额外评估显示，该优势高度依赖组织病理学模态的可用性，且当模态缺失时，我们的系统会落后于基线。我们的代码可在GitHub上获取，链接为this https URL。
英文摘要
Brain tumor diagnosis is a time-sensitive process in which patients may wait weeks for a finalized pathology report. This problem motivates automated systems that classify tumor subtype from multimodal inputs. This paper details the DS@GT ARC team's work for ImageCLEFmed MEDIQA-CORE 2026 Task~1, Brain Tumor Subtype Classification. The task evaluates three glioma classification problems: Level-1 Molecular Type, LGG vs HGG, and WHO Grade. We combine pre-extracted MRI (NeuroVFM) and histopathology (Prov-GigaPath) embeddings with free-text radiology reports. Our team explored two trimodal fusion architectures, two report encoders (RadBERT and Llama-3.1-8B-Instruct), and a biologically motivated post-processing stage. We achieve a mean macro-F1 of 0.801 under the Fully Multimodal condition, exceeding the organizers' baseline of 0.796 and ranking second among the teams whose code passed verification. Additional evaluation across modality-dropping conditions shows that this advantage depends heavily on the availability of the histopathology modality, and that our system falls behind the baseline when modalities are missing. Our code is available on GitHub at https://github.com/dsgt-arc/imageclef-mediqacore-2026.
CommentsCLEF 2026 Working Notes, 21 - 24 September 2026, Jena, Germany
