Towards continuous learning for glioma segmentation with elastic weight consolidation

Karin van Garderen, Sebastian van der Voort, Fatih Incekara, Marion Smits, Stefan Klein

Keywords: convolutional neural network, glioma, segmentation, continuous learning, magnetic resonance imaging

TL;DR: Evaluation of Elastic Weight Consolidation on catastrophic forgetting when re-training a segmentation network on a new domain.

Abstract: When finetuning a convolutional neural network (CNN) on data from a new domain, catastrophic forgetting will reduce performance on the original training data. Elastic Weight Consolidation (EWC) is a recent technique to prevent this, which we evaluated while training and re-training a CNN to segment glioma on two different datasets. The network was trained on the public BraTS dataset and finetuned on an in-house dataset with non-enhancing low-grade glioma. EWC was found to decrease catastrophic forgetting in this case, but was also found to restrict adaptation to the new domain.

Code Of Conduct: I have read and accept the code of conduct.

Remove If Rejected: Remove submission from public view if paper is rejected.