Joint Modelling Histology and Molecular Markers for Cancer Classification Journal: Medical Image Analysis
Xiaofei Wang, Hanyu Liu, Yupei Zhang, Boyang Zhao, Hao Duan, Wanming Hu, Yonggao Mou, Stephen Price, Chao Li · 2025
Read Joint Modelling Histology and Molecular Markers for Cancer Classification Journal: Medical Image Analysis by Xiaofei Wang, Hanyu Liu, Yupei Zhang, Boyang Zhao, Hao Duan, Wanming Hu, Yonggao Mou, Stephen Price, Chao Li (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-07979.txt.
Description
Wang and colleagues wrote in twenty twenty-five about a digital pathology method that reads tissue images and molecular markers together. Cancer classification increasingly needs both microscope features and molecular signals. Their method predicts both jointly and models how they interact. A multi-scale module handles information from cell-level to tissue-level views. An attention-based framework then predicts histology and markers at once. A graph network captures which markers tend to occur together.
A cross-modal module lets the two information types correct each other. Confidence-based losses and gradient balancing keep training stable. Tests focused on glioma, a brain tumor with mixed features. The method beat leading rivals at classifying tumors and markers. The authors say joint reading mirrors how pathologists now work. It could support precise oncology in clinics and research. Code is shared publicly for reuse.
Catalog record
| CREATOR | Xiaofei Wang, Hanyu Liu, Yupei Zhang, Boyang Zhao, Hao Duan, Wanming Hu, Yonggao Mou, Stephen Price, Chao Li |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 68 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-07979