面向病理全切片图像的无分布外检测训练方法
Training-Free Out-of-Distribution Detection for Pathology Whole-Slide Images
浏览论文内容
中文总结 AI 辅助
本文提出基于视觉-语言病理基础模型的无训练多模态OOD检测器ZIO，在14700余张病理WSI上优于40种现有OOD方法，为临床AI安全部署提供支撑。
中文摘要 AI 辅助
AI方法在医疗领域的安全部署需要可靠的防护机制，用于检测输入数据偏离训练分布的情况，确保模型仅在其专业范围内提供预测，否则弃权（不执行）。分布外（OOD）检测可提供此类防护，已在通用计算机视觉领域得到广泛研究，但在计算病理学领域仍发展不足，千兆像素级全切片图像（WSI）、疾病亚型间的细微差异及组织制备的变异性，给传统OOD方法带来独特挑战。本文提出ZIO，一种面向病理WSI的无训练多模态OOD检测器，利用视觉-语言病理基础模型（FM），构建分布内类别的文本和视觉原型，并通过原型收缩机制整合二者互补信息以推导OOD分数，提供适用于切片级和补丁级FM的ZIO公式。在涵盖罕见疾病和近OOD设置的多种临床相关域偏移场景中评估ZIO，对来自5个独立联盟的14700余张WSI的广泛评估显示，ZIO始终优于单模态原型及40种最先进的OOD方法。这些结果证明了多模态表示在OOD检测中的优势，为临床实践中更安全的AI部署铺平了道路。
英文摘要
Safe deployment of AI methods in medicine requires robust guardrails that detect when input data deviate from the training distribution to ensure that models provide predictions only within their scope of expertise and abstain otherwise. Out-of-distribution (OOD) detection can provide such safeguards and is extensively studied in general computer vision. Yet, it remains underdeveloped in computational pathology, where gigapixel whole-slide images (WSIs), subtle differences between disease subtypes, and variability in tissue preparation pose unique challenges for conventional OOD methods. We propose ZIO, a training-free, multimodal OOD detector for pathology WSIs that leverages vision--language pathology foundation models (FMs). ZIO constructs text and visual prototypes of in-distribution classes and integrates their complementary information through a prototype shrinkage mechanism to derive OOD scores. We provide the ZIO formulation for both slide- and patch-level FMs. We evaluate ZIO across diverse clinically relevant domain shifts, including rare diseases and near-OOD settings. Extensive evaluation of over 14,700 WSIs from five independent consortia shows that ZIO consistently outperforms both unimodal prototypes and 40 state-of-the-art OOD methods. These results demonstrate the benefits of multimodal representation for OOD detection and pave the way towards safer AI deployment in clinical practice.
