Training-Free Out-of-Distribution Detection for Pathology Whole-Slide Images

2026-08-02

AI Summary

This work addresses the challenges of out-of-distribution (OOD) detection in whole-slide images (WSIs), which arise from their gigapixel scale, subtle differences among disease subtypes, and significant variations in sample preparation. To tackle these issues, the authors propose ZIO—the first training-free, multimodal OOD detection method for computational pathology. ZIO leverages a vision–language foundation model to construct both textual and visual prototypes and introduces a prototype contraction mechanism to fuse them into a unified OOD score, enabling both slide-level and patch-level inference. Evaluated across five consortia with over 14,700 WSIs, ZIO substantially outperforms unimodal approaches and 40 state-of-the-art OOD detection algorithms, demonstrating exceptional performance in identifying rare diseases and near-OOD samples.

Abstract

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.

Problem

• out-of-distribution detection
• computational pathology
• whole-slide images
• domain shift
• AI safety

Innovation

• training-free
• multimodal OOD detection
• vision-language foundation models
• prototype shrinkage
• whole-slide images