Few-Shot Continual Learning for 3D Brain MRI with Frozen Foundation Models
This work provides a practical solution for medical imaging practitioners who need to adapt large 3D medical foundation models to new tasks sequentially with limited data, while maintaining performance on all learned tasks.
This paper tackles few-shot continual learning for 3D brain MRI by using a frozen pretrained backbone combined with task-specific Low-Rank Adaptation (LoRA) modules. This approach achieves a balanced performance across tumor segmentation (Dice 0.62) and brain age estimation (MAE 0.16) with zero forgetting and less than 0.1% trainable parameters per task.
Foundation models pretrained on large-scale 3D medical imaging data face challenges when adapted to multiple downstream tasks under continual learning with limited labeled data. We address few-shot continual learning for 3D brain MRI by combining a frozen pretrained backbone with task-specific Low-Rank Adaptation (LoRA) modules. Tasks arrive sequentially -- tumor segmentation (BraTS) and brain age estimation (IXI) -- with no replay of previous task data. Each task receives a dedicated LoRA adapter; only the adapter and task-specific head are trained while the backbone remains frozen, thereby eliminating catastrophic forgetting by design (BWT=0). In continual learning, sequential full fine-tuning suffers severe forgetting (T1 Dice drops from 0.80 to 0.16 after T2), while sequential linear probing achieves strong T1 (Dice 0.79) but fails on T2 (MAE 1.45). Our LoRA approach achieves the best balanced performance across both tasks: T1 Dice 0.62$\pm$0.07, T2 MAE 0.16$\pm$0.05, with zero forgetting and $<$0.1\% trainable parameters per task, though with noted systematic age underestimation in T2 (Wilcoxon $p<0.001$). Frozen foundation models with task-specific LoRA adapters thus offer a practical solution when both tasks must be maintained under few-shot continual learning.