HyperCT: Low-Rank Hypernet for Unified Chest CT Analysis
This work addresses the need for a unified, parameter-efficient solution for holistic patient assessment from non-contrast chest CTs, though it is incremental as it builds on existing hypernetwork and low-rank adaptation methods.
The authors tackled the problem of suboptimal performance in multi-task learning for chest CT analysis by proposing HyperCT, a framework that dynamically adapts a Vision Transformer backbone using a hypernetwork with low-rank adaptation, which outperformed strong baselines on a large-scale dataset of radiological and cardiological tasks.
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.