Integrating deep learning with multimodal MRI habitat radiomics: toward personalized prediction of risk stratification and androgen deprivation therapy outcomes in prostate cancer
Multi-model integration of habitat radiomics and 3D Vision Transformer achieves superior prediction for ADT response compared to conventional methods. Habitat radiomics outperforms traditional radiomics in Gleason score stratification. SHAP analysis provides clinical interpretability, identifying key model linked to ADT outcomes for actionable insights.
Authors
- Qidong Wang (ORCID: https://orcid.org/0000-0003-2082-8457)
- Lei Zhao (ORCID: https://orcid.org/0000-0003-3416-2254)
- Yunfeng Zhang (ORCID: https://orcid.org/0000-0001-6627-2541)
- Han He
- Fenghai Zhou (ORCID: https://orcid.org/0009-0004-8358-1943)
- Wenbo Zhang
- Rujing Wang (ORCID: https://orcid.org/0000-0002-0115-1156)
- Hongde Hu
- Chao Wang
- Jie Yang
- Jia Wang
- Wanbin He
- Chuan Zhou
Institutions
- University of Electronic Science and Technology of China (CN)
- Gansu University of Traditional Chinese Medicine (CN)
- Gansu Provincial Hospital (CN)
- Lanzhou University Second Hospital (CN)
- First Hospital of Lanzhou University (CN)
- Lanzhou University (CN)
Publication Details
- Journal
- Insights into Imaging
- Published
- 2026-01-26
- DOI
- https://doi.org/10.1186/s13244-026-02205-8
- Primary Topic
- Prostate Cancer Diagnosis and Treatment
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Lanzhou University
- Science and Technology Program of Gansu Province