Medical hierarchical image classification via dual-geometry image–text learning
CL achieves an average accuracy improvement of 7% across all three datasets at the fine-grained level, with similarly consistent gains observed when integrated with other backbone models. The source code is publicly available at https://github.com/MCPathology/H2CL.
Authors
- Arcot Sowmya (ORCID: https://orcid.org/0000-0001-9236-5063)
- Zhen Yu
- Yang Song
- Erik Meijering (ORCID: https://orcid.org/0000-0001-8015-8358)
- Zongyuan Ge
- Lei Fan (ORCID: https://orcid.org/0000-0001-9472-7152)
Institutions
- UNSW Sydney (AU)
- Monash University (AU)
Publication Details
- Journal
- Medical Image Analysis
- Published
- 2026-04-30
- DOI
- https://doi.org/10.1016/j.media.2026.104120
- Primary Topic
- AI in cancer detection
- Type
- article
- Field-Weighted Citation Impact
- 0.00