文献速递医学人工智能与信息学基础与转化研究

面向多样医疗应用的开放视觉语言模型

An open vision-language model for diverse medical applications.
PubMed 文献 · Nat Med · 2026-10-06

研究提出基于Gemma 3的医学视觉语言基础模型集MedGemma及医学调优视觉编码器MedSigLIP。MedGemma在图像与文本的医学理解与推理上超过同规模生成模型,在分布外任务中医学图像问答提升2.6-10%、胸片表现分类提升15.5-18.1%、智能体评估提升10.8%。在训练数据有限时,微调MedGemma比微调基础Gemma 3更有效。

要点

  • MedGemma在分布外医学图像问答提升2.6-10%。
  • 胸片表现分类提升15.5-18.1%,智能体评估提升10.8%。
  • 数据有限时微调MedGemma优于微调基础Gemma 3。

核心结论

  • MedGemma集合提供较强的医学图像与文本基础能力,有望加速医学研究与下游应用开发。

研究设计与样本

  • 研究构建基于Gemma 3的医学视觉语言基础模型集MedGemma及医学调优视觉编码器MedSigLIP,并与同规模生成模型及基础模型比较。

主要发现

  • MedGemma在图像和文本及多个医学影像领域的医学理解与推理超过同规模生成模型,同时保持Gemma基础模型的通用能力。
  • 分布外任务中,医学图像问答提升2.6-10%,胸片表现分类提升15.5-18.1%,智能体评估提升10.8%。
  • 微调MedGemma在医学任务上比微调基础Gemma 3更有效,尤其在训练数据有限时。
  • MedSigLIP作为编码器性能与许多专用医学图像编码器相当或更优。

临床怎么用

  • 在训练数据有限时,可优先微调MedGemma而非基础Gemma 3以提升医学任务表现。
  • MedSigLIP可作为医学视觉编码器用于图像理解任务。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

Artificial intelligence has high potential for impact in healthcare applications, but its training and deployment are challenging due to diverse data, a complex spectrum of possible tasks and important privacy needs. High-performing foundation models that enable data-efficient fine-tuning for diverse downstream tasks can meaningfully accelerate development in this domain. We introduce MedGemma, a collection of medical vision-language foundation models based on Gemma 3. MedGemma demonstrates advanced medical understanding and reasoning across images and text and multiple medical imaging domains, exceeding the performance of similarly sized generative models while maintaining the general capabilities of the Gemma base models. For out-of-distribution tasks, MedGemma achieves improvements of 2.6-10% in medical image question answering, 15.5-18.1% in chest X-ray finding classification and 10.8% in agentic evaluations compared with the base models. Our results show that fine-tuning MedGemma can be more effective than fine-tuning the base Gemma 3 model for medical tasks, particularly in the setting of limited training data. We additionally introduce MedSigLIP, a medically tuned vision encoder derived from SigLIP. MedSigLIP powers the visual understanding capabilities of MedGemma and, as an encoder, achieves performance comparable to or better than that of many specialized medical image encoders. Taken together, the MedGemma collection provides a strong foundation of medical image and text capabilities, with the potential to accelerate medical research and the development of downstream applications.

来源:Nat Med,查看原文。版权归原作者所有。

引用

  1. Sellergren A, Kazemzadeh S, Mahvar F, et al. An open vision-language model for diverse medical applications. Nat Med. 2026. PMID: 42839104. doi:10.1038/s41591-026-04626-w

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。