Importance Driven Continual Learning for Segmentation Across Domains
This addresses the problem of retaining prior knowledge in medical applications where sensitive data cannot be stored, but it is incremental as it adapts an existing importance-driven method.
The paper tackles catastrophic forgetting in neural networks for medical image segmentation across domains by proposing a continual learning approach with learning rate regularization, demonstrating that restricting adaptation of important parameters reduces forgetting.
The ability of neural networks to continuously learn and adapt to new tasks while retaining prior knowledge is crucial for many applications. However, current neural networks tend to forget previously learned tasks when trained on new ones, i.e., they suffer from Catastrophic Forgetting (CF). The objective of Continual Learning (CL) is to alleviate this problem, which is particularly relevant for medical applications, where it may not be feasible to store and access previously used sensitive patient data. In this work, we propose a Continual Learning approach for brain segmentation, where a single network is consecutively trained on samples from different domains. We build upon an importance driven approach and adapt it for medical image segmentation. Particularly, we introduce learning rate regularization to prevent the loss of the network's knowledge. Our results demonstrate that directly restricting the adaptation of important network parameters clearly reduces Catastrophic Forgetting for segmentation across domains.