HyperCT: Low-Rank Hypernet for Unified Chest CT Analysis
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908935058882560 |
|---|---|
| author | Liu, Fengbei Kwak, Sunwoo Phung, Hao Nizam, Nusrat Binta Richter, Ilan Uriel, Nir Averbuch-Elor, Hadar Estrin, Daborah Sabuncu, Mert R. |
| author_facet | Liu, Fengbei Kwak, Sunwoo Phung, Hao Nizam, Nusrat Binta Richter, Ilan Uriel, Nir Averbuch-Elor, Hadar Estrin, Daborah Sabuncu, Mert R. |
| contents | Non-contrast chest CTs offer a rich opportunity for both conventional pulmonary and opportunistic extra-pulmonary screening. While Multi-Task Learning (MTL) can unify these diverse tasks, standard hard-parameter sharing approaches are often suboptimal for modeling distinct pathologies. We propose HyperCT, a framework that dynamically adapts a Vision Transformer backbone via a Hypernetwork. To ensure computational efficiency, we integrate Low-Rank Adaptation (LoRA), allowing the model to regress task-specific low-rank weight updates rather than full parameters. Validated on a large-scale dataset of radiological and cardiological tasks, \method{} outperforms various strong baselines, offering a unified, parameter-efficient solution for holistic patient assessment. Our code is available at https://github.com/lfb-1/HyperCT. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_03224 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | HyperCT: Low-Rank Hypernet for Unified Chest CT Analysis Liu, Fengbei Kwak, Sunwoo Phung, Hao Nizam, Nusrat Binta Richter, Ilan Uriel, Nir Averbuch-Elor, Hadar Estrin, Daborah Sabuncu, Mert R. Image and Video Processing Computer Vision and Pattern Recognition Non-contrast chest CTs offer a rich opportunity for both conventional pulmonary and opportunistic extra-pulmonary screening. While Multi-Task Learning (MTL) can unify these diverse tasks, standard hard-parameter sharing approaches are often suboptimal for modeling distinct pathologies. We propose HyperCT, a framework that dynamically adapts a Vision Transformer backbone via a Hypernetwork. To ensure computational efficiency, we integrate Low-Rank Adaptation (LoRA), allowing the model to regress task-specific low-rank weight updates rather than full parameters. Validated on a large-scale dataset of radiological and cardiological tasks, \method{} outperforms various strong baselines, offering a unified, parameter-efficient solution for holistic patient assessment. Our code is available at https://github.com/lfb-1/HyperCT. |
| title | HyperCT: Low-Rank Hypernet for Unified Chest CT Analysis |
| topic | Image and Video Processing Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.03224 |