An interpretable model based on weakly supervised learning for uterine smooth muscle tumor diagnosis: A multi-center study
Pathology - Research and Practice · 2026
DOI: 10.1016/j.prp.2025.156337 ↗Kurz gefasst
Multizentrische Studie in der Zeitschrift Pathology - Research and Practice, 2026. Verfasst von Xiaoxi Wang und 19 weiteren Autor:innen. Beteiligte Einrichtung: China-Japan Friendship Hospital (Institute of Clinical Medical Sciences), Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing…. Hauptthemen laut MEDLINE: Uterine Neoplasms, Smooth Muscle Tumor, Supervised Machine Learning, Leiomyoma. Die Arbeit nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Chinese Academy of Medical Sciences, National Natural Science Foundation of China.
Abstract (Originalsprache)
Uterine smooth muscle tumors (USMTs) are the most common tumors of the female reproductive system, but remain diagnostically challenging due to morphological overlap among leiomyosarcoma (LMS), various leiomyoma (LM) subtypes, and smooth muscle tumors of uncertain malignant potential (STUMP). This study aimed to develop a weakly supervised artificial intelligence (AI) model for classifying USMTs as benign or malignant using only slide-level labels. A multi-center dataset comprising 94 LMS cases (408 whole-slide images, WSIs) and 634 benign cases (1389 WSIs) was used for model training and internal testing, with an independent external test set including 27 LMS cases (54 WSIs) and 90 benign cases (248 WSIs). The CAMEL2-based model achieved excellent diagnostic performance, with AUCs of 0.9976 and 0.9889 on the internal and external test sets, respectively, and accuracies exceeding 0.97. Heatmaps frequently highlighted regions enriched for key pathological features of LMS, while benign tissues were consistently assigned low-suspicion regions. In STUMP cases, heatmaps emphasized morphologically suspicious regions that often overlapped with areas identified by pathologists, supporting their potential as decision-support visualizations. In the AI-human collaboration study, model assistance was associated with improved diagnostic accuracy and reduced diagnostic time. This represents the first weakly supervised learning model for USMT diagnosis, achieving high accuracy and interpretability with minimal annotation requirements.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- China-Japan Friendship Hospital (Institute of Clinical Medical Sciences), Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China.
- School of Mathematical Sciences, University of Chinese Academy of Sciences, Beijing, China.
- Department of Pathology, Suining Central Hospital, Suining, Sichuan, China.
- Department of Pathology, West China Second University Hospital, Sichuan University, Chengdu, China; Laboratory of Molecular Translational Medicine, Center for Translational Medicine, Key Laboratory of Birth Defects & Re…
- Department of Pathology, Clinical Laboratory Medicine Center, Inner Mongolia People's Hospital, Hohhot, Inner Mongolia, China.
- Thorough Lab, Thorough Future, Beijing, China.
Förderung
- Chinese Academy of Medical Sciences
- National Natural Science Foundation of China
Zitieren
APA
Wang, X., Shen, X., Yang, M., Yang, L., He, Y., Ren, Y., Wang, S., Wu, D., Guo, Y., Niu, Y., Peng, K., Wang, Z., Wu, J., Li, J., Wang, W., Wang, F., Cui, L., Wang, W., Liu, C., & Zhong, D. (2026). An interpretable model based on weakly supervised learning for uterine smooth muscle tumor diagnosis: A multi-center study. Pathology - Research and Practice. https://doi.org/10.1016/j.prp.2025.156337
Vancouver
Wang X, Shen X, Yang M, Yang L, He Y, Ren Y, et al. An interpretable model based on weakly supervised learning for uterine smooth muscle tumor diagnosis: A multi-center study. Pathology - Research and Practice. 2026. doi:10.1016/j.prp.2025.156337
BibTeX
@article{Wang2026,
author = {Wang, Xiaoxi and Shen, Xiaochen and Yang, Moxuan and Yang, Ling and He, Ying and Ren, Yanni and Wang, Shuhao and Wu, Dayi and Guo, Yuanyuan and Niu, Yiru and Peng, Kaiyue and Wang, Zhihui and Wu, Jingpeng and Li, Jiacheng and Wang, Wei and Wang, Feng and Cui, Lijuan and Wang, Wei and Liu, Congrong and Zhong, Dingrong},
title = {An interpretable model based on weakly supervised learning for uterine smooth muscle tumor diagnosis: A multi-center study},
journal = {Pathology - Research and Practice},
year = {2026},
doi = {10.1016/j.prp.2025.156337}
}
EUVIMED