_version_ 1866908395847548928
author Zhu, Lianghui
Ling, Xitong
Ouyang, Minxi
Liu, Xiaoping
Guan, Tian
Fu, Mingxi
Cheng, Zhiqiang
Fu, Fanglei
Zeng, Maomao
Liu, Liming
Duan, Song
Huang, Qiang
Xiao, Ying
Li, Jianming
Lu, Shanming
Piao, Zhenghua
Zhu, Mingxi
Jin, Yibo
Xu, Shan
He, Qiming
Wang, Yizhi
Cheng, Junru
Wang, Xuanyu
Xie, Luxi
Li, Houqiang
Tian, Sufang
He, Yonghong
author_facet Zhu, Lianghui
Ling, Xitong
Ouyang, Minxi
Liu, Xiaoping
Guan, Tian
Fu, Mingxi
Cheng, Zhiqiang
Fu, Fanglei
Zeng, Maomao
Liu, Liming
Duan, Song
Huang, Qiang
Xiao, Ying
Li, Jianming
Lu, Shanming
Piao, Zhenghua
Zhu, Mingxi
Jin, Yibo
Xu, Shan
He, Qiming
Wang, Yizhi
Cheng, Junru
Wang, Xuanyu
Xie, Luxi
Li, Houqiang
Tian, Sufang
He, Yonghong
contents Gastrointestinal (GI) diseases represent a clinically significant burden, necessitating precise diagnostic approaches to optimize patient outcomes. Conventional histopathological diagnosis suffers from limited reproducibility and diagnostic variability. To overcome these limitations, we develop Digepath, a specialized foundation model for GI pathology. Our framework introduces a dual-phase iterative optimization strategy combining pretraining with fine-screening, specifically designed to address the detection of sparsely distributed lesion areas in whole-slide images. Digepath is pretrained on over 353 million multi-scale images from 210,043 H&E-stained slides of GI diseases. It attains state-of-the-art performance on 33 out of 34 tasks related to GI pathology, including pathological diagnosis, protein expression status prediction, gene mutation prediction, and prognosis evaluation. We further translate the intelligent screening module for early GI cancer and achieve near-perfect 99.70% sensitivity across nine independent medical institutions. This work not only advances AI-driven precision pathology for GI diseases but also bridge critical gaps in histopathological practice.
format Preprint
id arxiv_https___arxiv_org_abs_2505_21928
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Subspecialty-Specific Foundation Model for Intelligent Gastrointestinal Pathology
Zhu, Lianghui
Ling, Xitong
Ouyang, Minxi
Liu, Xiaoping
Guan, Tian
Fu, Mingxi
Cheng, Zhiqiang
Fu, Fanglei
Zeng, Maomao
Liu, Liming
Duan, Song
Huang, Qiang
Xiao, Ying
Li, Jianming
Lu, Shanming
Piao, Zhenghua
Zhu, Mingxi
Jin, Yibo
Xu, Shan
He, Qiming
Wang, Yizhi
Cheng, Junru
Wang, Xuanyu
Xie, Luxi
Li, Houqiang
Tian, Sufang
He, Yonghong
Image and Video Processing
Artificial Intelligence
Computer Vision and Pattern Recognition
Machine Learning
Gastrointestinal (GI) diseases represent a clinically significant burden, necessitating precise diagnostic approaches to optimize patient outcomes. Conventional histopathological diagnosis suffers from limited reproducibility and diagnostic variability. To overcome these limitations, we develop Digepath, a specialized foundation model for GI pathology. Our framework introduces a dual-phase iterative optimization strategy combining pretraining with fine-screening, specifically designed to address the detection of sparsely distributed lesion areas in whole-slide images. Digepath is pretrained on over 353 million multi-scale images from 210,043 H&E-stained slides of GI diseases. It attains state-of-the-art performance on 33 out of 34 tasks related to GI pathology, including pathological diagnosis, protein expression status prediction, gene mutation prediction, and prognosis evaluation. We further translate the intelligent screening module for early GI cancer and achieve near-perfect 99.70% sensitivity across nine independent medical institutions. This work not only advances AI-driven precision pathology for GI diseases but also bridge critical gaps in histopathological practice.
title Subspecialty-Specific Foundation Model for Intelligent Gastrointestinal Pathology
topic Image and Video Processing
Artificial Intelligence
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2505.21928