Benchmarking Vision-Language Models for Automated Pathology Diagnosis and Report Generation

2026-09-01

AI Summary

为解决WSI病理报告生成数据稀缺和复杂性问题，构建了泛亚洲WSI-报告数据集，并通过MICCAI挑战赛评估多种模型方法。

Abstract

The rapid advancement of vision-language models (VLMs) has accelerated progress in computational pathology; however, whole-slide image (WSI)-based pathology report generation remains limited by the scarcity of large-scale WSI--report datasets and the complexity of mapping spatially distributed visual patterns to structured clinical text. To address this, we introduce a clinically curated Pan-Asia WSI--report dataset of approximately 10,500 pairs from five institutions and establish the REG 2025 benchmark through a MICCAI challenge for systematic evaluation of multimodal models. We analyze submitted methods spanning pretrained VLMs, multiple-instance learning frameworks, hierarchical expert models, retrieval-augmented generation, and cross-modal Transformers. Rather than indicating that VLM use alone was sufficient for superior performance, the results suggest that top-performing methods benefited from structured report representations, hierarchical diagnostic decomposition, and effective multimodal grounding. We identify key limitations, including instability in quantitative attribute estimation (e.g., numeric hallucination) and a tendency toward diagnostic overspecification, with some errors resembling known diagnostic pitfalls in routine pathology. These findings establish REG 2025 as a benchmark for evaluating WSI-based structured report generation and vision-language understanding in computational pathology, providing insights for the design of clinically grounded multimodal pathology models.

Problem

• vision-language models
• whole-slide image
• pathology report generation
• multimodal models

Innovation

• Vision-Language Models
• Whole-Slide Image
• Structured Report Representations
• Hierarchical Diagnostic Decomposition
• Multimodal Grounding