paper-with-me

홈 › Papers

Do Pathology Vision-Language Models Truly See Pathology?

2026-07-23 · Chengyang Zhang, Wenchuan Zhang, Bo Li, Xinyu Liu, Jiaming Yang, Mengran Li, Chenxun Deng, Jie Chen, Yang Zhang, Wei Ju, Yuhao Yi, Hong Bu, Jiancheng Lv arxiv

Pathology vision-language models (VLMs) have recently progressed rapidly and are commonly evaluated by answer accuracy on pathology VQA benchmarks. However, we dig into current evaluations and identify three overlooked issues: 1) Visual evidence is not always necessary. For instance, Gemini-3-Pro achieves 53.5% average accuracy across 5 VQA benchmarks without any visual input. 2) Domain training can improve accuracy without proportional gains in visual binding. Compared with Qwen2.5-VL-7B, Patho-R1-7B exhibits a 5.8-point lower multimodal gain and a 3.7-point lower attention IoU. 3) Entity-level attention is diffuse and weakly query-specific. On PathVG, attention maps remain highly correlated across different entity queries. These issues can lead to substantial misjudgments of pathology VLMs' actual multimodal capabilities. To this end, we present PathBind, a benchmark comprising 2,600 samples: PathBind-VQA with 1,500 questions across six dimensions, PathBind-PTA with 600 questions from a private pathology teaching atlas, and PathBind-Grounding with 500 expert-curated region-level samples. Each component undergoes task-specific automated filtering and expert review to reduce textual shortcuts and improve entity-region correspondence. We evaluate 18 representative VLMs on VQA samples of PathBind and five existing pathology VQA benchmarks, and further evaluate 10 VLMs on PathBind-Grounding and PathVG. Results show that current pathology VLMs still exhibit a substantial gap between answer-side performance and visual-semantic binding.

📄 PDF Abstract BibTeX arXiv:2607.21065

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Benchmarking foundation models as feature extractors for weakly-supervised computational pathology

2024-08-28 · Peter Neidlinger, Omar S. M. El Nahhas, Hannah Sophie Muti, Tim Lenz 외

Advancements in artificial intelligence have driven the development of numerous pathology foundation models capable of extracting clinically relevant information. However, there is currently limited literature independen…

BenchmarkingDiversity

Towards a text-based quantitative and explainable histopathology image analysis

2024-07-10 · Anh Tien Nguyen, Trinh Thi Le Vuong, Jin Tae Kwak

Recently, vision-language pre-trained models have emerged in computational pathology. Previous works generally focused on the alignment of image-text pairs via the contrastive pre-training paradigm. Such pre-trained mode…

image-classificationImage ClassificationImage to textImage-to-Text Retrieval+5

Normal and Abnormal Pathology Knowledge-Augmented Vision-Language Model for Anomaly Detection in Pathology Images

2025-08-21 · Jinsol Song, Jiamu Wang, Anh Tien Nguyen, Keunho Byeon 외 arxiv

Anomaly detection in computational pathology aims to identify rare and scarce anomalies where disease-related data are often limited or missing. Existing anomaly detection methods, primarily designed for industrial setti…

Anomaly Detection

A Foundational Multimodal Vision Language AI Assistant for Human Pathology

2023-12-13 · Ming Y. Lu, Bowen Chen, Drew F. K. Williamson, Richard J. Chen 외

The field of computational pathology has witnessed remarkable progress in the development of both task-specific predictive models and task-agnostic self-supervised vision encoders. However, despite the explosive growth o…

Decision MakingDiagnosticLanguage ModellingLarge Language Model+1

CPLIP: Zero-Shot Learning for Histopathology with Comprehensive Vision-Language Alignment

2024-06-07 · CVPR 2024 1 · Sajid Javed, Arif Mahmood, Iyyakutti Iyappan Ganapathi, Fayaz Ali Dharejo 외

This paper proposes Comprehensive Pathology Language Image Pre-training (CPLIP), a new unsupervised technique designed to enhance the alignment of images and text in histopathology for tasks such as classification and se…

Contrastive LearningZero-Shot Learning