Saved in:
Bibliographic Details
Main Authors: Ge, Jiusong, Zhan, Yingkang, Zhao, Wenjie, Zhang, Di, Wang, Ke, Liu, Jiashuai, Yang, Chunze, Li, Chengzu, Zhang, Jian, Dong, Yuxin, Zhang, Ni, Liu, Qidong, Crispin-Ortuzar, Mireia, Fu, Huazhu, Li, Chen, Gao, Zeyu
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.19491
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917527672586240
author Ge, Jiusong
Zhan, Yingkang
Zhao, Wenjie
Zhang, Di
Wang, Ke
Liu, Jiashuai
Yang, Chunze
Li, Chengzu
Zhang, Jian
Dong, Yuxin
Zhang, Ni
Liu, Qidong
Crispin-Ortuzar, Mireia
Fu, Huazhu
Li, Chen
Gao, Zeyu
author_facet Ge, Jiusong
Zhan, Yingkang
Zhao, Wenjie
Zhang, Di
Wang, Ke
Liu, Jiashuai
Yang, Chunze
Li, Chengzu
Zhang, Jian
Dong, Yuxin
Zhang, Ni
Liu, Qidong
Crispin-Ortuzar, Mireia
Fu, Huazhu
Li, Chen
Gao, Zeyu
contents Traditional whole slide image (WSI) analysis methods typically rely on the multiple instance learning (MIL) paradigm, which extracts patch-level features at high magnification and aggregates them for slide-level prediction. However, such exhaustive patch-level processing is computationally expensive, severely limiting the efficiency and scalability of WSI analysis. To address this challenge, we propose PathCTM (a Pathology-oriented Continuous Thought Model) that enables token-efficient scale-space continuous reasoning for gigapixel WSIs. PathCTM formulates diagnostic inference as a dynamic sequential information pursuit. It progressively transitions from low-magnification global to high-magnification local inspection, and adaptively terminates inference when sufficient evidence is gathered to effectively bound decision uncertainty. Specifically, it uses conditional computation for dynamic scale switching with attention-guided region pruning, coupled with confidence-aware early stopping. Extensive experiments demonstrate that, compared with standard MIL-based methods, PathCTM reduces the number of required image patches by 95.95% and shortens inference time by approximately 95.62%, while maintaining AUC without degradation. Code is available at https://github.com/JSGe-AI/PathCTM.
format Preprint
id arxiv_https___arxiv_org_abs_2605_19491
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Thinking in Scales: Accelerating Gigapixel Pathology Image Analysis via Adaptive Continuous Reasoning
Ge, Jiusong
Zhan, Yingkang
Zhao, Wenjie
Zhang, Di
Wang, Ke
Liu, Jiashuai
Yang, Chunze
Li, Chengzu
Zhang, Jian
Dong, Yuxin
Zhang, Ni
Liu, Qidong
Crispin-Ortuzar, Mireia
Fu, Huazhu
Li, Chen
Gao, Zeyu
Computer Vision and Pattern Recognition
Traditional whole slide image (WSI) analysis methods typically rely on the multiple instance learning (MIL) paradigm, which extracts patch-level features at high magnification and aggregates them for slide-level prediction. However, such exhaustive patch-level processing is computationally expensive, severely limiting the efficiency and scalability of WSI analysis. To address this challenge, we propose PathCTM (a Pathology-oriented Continuous Thought Model) that enables token-efficient scale-space continuous reasoning for gigapixel WSIs. PathCTM formulates diagnostic inference as a dynamic sequential information pursuit. It progressively transitions from low-magnification global to high-magnification local inspection, and adaptively terminates inference when sufficient evidence is gathered to effectively bound decision uncertainty. Specifically, it uses conditional computation for dynamic scale switching with attention-guided region pruning, coupled with confidence-aware early stopping. Extensive experiments demonstrate that, compared with standard MIL-based methods, PathCTM reduces the number of required image patches by 95.95% and shortens inference time by approximately 95.62%, while maintaining AUC without degradation. Code is available at https://github.com/JSGe-AI/PathCTM.
title Thinking in Scales: Accelerating Gigapixel Pathology Image Analysis via Adaptive Continuous Reasoning
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.19491