ModalSurv: Investigating opportunities and limitations of multimodal deep survival learning in prostate and bladder cancer
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918253451804672 |
|---|---|
| author | Wahab, Noorul Alzaid, Ethar Lv, Jiaqi Minhas, Fayyaz Shephard, Adam Raza, Shan E Ahmed |
| author_facet | Wahab, Noorul Alzaid, Ethar Lv, Jiaqi Minhas, Fayyaz Shephard, Adam Raza, Shan E Ahmed |
| contents | Accurate survival prediction is essential for personalised cancer treatment. We propose ModalSurv, a multimodal deep survival framework integrating clinical, MRI, histopathology, and RNA-sequencing data via modality-specific projections and cross-attention fusion. On the CHIMERA Grand Challenge datasets, ModalSurv achieved a C-index of 0.7402 (1st) for prostate and 0.5740 (5th) for bladder cancer. Notably, clinical features alone outperformed multimodal models on external tests, highlighting challenges of limited multimodal alignment and potential overfitting. Local validation showed multimodal gains but limited generalisation. ModalSurv provides a systematic evaluation of multimodal survival modelling, underscoring both its promise and current limitations for scalable, generalisable cancer prognosis. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_05037 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ModalSurv: Investigating opportunities and limitations of multimodal deep survival learning in prostate and bladder cancer Wahab, Noorul Alzaid, Ethar Lv, Jiaqi Minhas, Fayyaz Shephard, Adam Raza, Shan E Ahmed Machine Learning Accurate survival prediction is essential for personalised cancer treatment. We propose ModalSurv, a multimodal deep survival framework integrating clinical, MRI, histopathology, and RNA-sequencing data via modality-specific projections and cross-attention fusion. On the CHIMERA Grand Challenge datasets, ModalSurv achieved a C-index of 0.7402 (1st) for prostate and 0.5740 (5th) for bladder cancer. Notably, clinical features alone outperformed multimodal models on external tests, highlighting challenges of limited multimodal alignment and potential overfitting. Local validation showed multimodal gains but limited generalisation. ModalSurv provides a systematic evaluation of multimodal survival modelling, underscoring both its promise and current limitations for scalable, generalisable cancer prognosis. |
| title | ModalSurv: Investigating opportunities and limitations of multimodal deep survival learning in prostate and bladder cancer |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2509.05037 |