Single Domain Generalization for Multimodal Cross-Cancer Prognosis via Dirac Rebalancer and Distribution Entanglement
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916924676374528 |
|---|---|
| author | Jiang, Jia-Xuan Liu, Jiashuai Wu, Hongtao Wu, Yifeng Wang, Zhong Bi, Qi Zheng, Yefeng |
| author_facet | Jiang, Jia-Xuan Liu, Jiashuai Wu, Hongtao Wu, Yifeng Wang, Zhong Bi, Qi Zheng, Yefeng |
| contents | Deep learning has shown remarkable performance in integrating multimodal data for survival prediction. However, existing multimodal methods mainly focus on single cancer types and overlook the challenge of generalization across cancers. In this work, we are the first to reveal that multimodal prognosis models often generalize worse than unimodal ones in cross-cancer scenarios, despite the critical need for such robustness in clinical practice. To address this, we propose a new task: Cross-Cancer Single Domain Generalization for Multimodal Prognosis, which evaluates whether models trained on a single cancer type can generalize to unseen cancers. We identify two key challenges: degraded features from weaker modalities and ineffective multimodal integration. To tackle these, we introduce two plug-and-play modules: Sparse Dirac Information Rebalancer (SDIR) and Cancer-aware Distribution Entanglement (CADE). SDIR mitigates the dominance of strong features by applying Bernoulli-based sparsification and Dirac-inspired stabilization to enhance weaker modality signals. CADE, designed to synthesize the target domain distribution, fuses local morphological cues and global gene expression in latent space. Experiments on a four-cancer-type benchmark demonstrate superior generalization, laying the foundation for practical, robust cross-cancer multimodal prognosis. Code is available at https://github.com/HopkinsKwong/MCCSDG |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_08340 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Single Domain Generalization for Multimodal Cross-Cancer Prognosis via Dirac Rebalancer and Distribution Entanglement Jiang, Jia-Xuan Liu, Jiashuai Wu, Hongtao Wu, Yifeng Wang, Zhong Bi, Qi Zheng, Yefeng Computer Vision and Pattern Recognition Artificial Intelligence Deep learning has shown remarkable performance in integrating multimodal data for survival prediction. However, existing multimodal methods mainly focus on single cancer types and overlook the challenge of generalization across cancers. In this work, we are the first to reveal that multimodal prognosis models often generalize worse than unimodal ones in cross-cancer scenarios, despite the critical need for such robustness in clinical practice. To address this, we propose a new task: Cross-Cancer Single Domain Generalization for Multimodal Prognosis, which evaluates whether models trained on a single cancer type can generalize to unseen cancers. We identify two key challenges: degraded features from weaker modalities and ineffective multimodal integration. To tackle these, we introduce two plug-and-play modules: Sparse Dirac Information Rebalancer (SDIR) and Cancer-aware Distribution Entanglement (CADE). SDIR mitigates the dominance of strong features by applying Bernoulli-based sparsification and Dirac-inspired stabilization to enhance weaker modality signals. CADE, designed to synthesize the target domain distribution, fuses local morphological cues and global gene expression in latent space. Experiments on a four-cancer-type benchmark demonstrate superior generalization, laying the foundation for practical, robust cross-cancer multimodal prognosis. Code is available at https://github.com/HopkinsKwong/MCCSDG |
| title | Single Domain Generalization for Multimodal Cross-Cancer Prognosis via Dirac Rebalancer and Distribution Entanglement |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2507.08340 |