Towards continual learning in medical imaging

Abstract

This work investigates continual learning of two segmentation tasks in brain MRI with neural networks. To explore in this context the capabilities of current methods for countering catastrophic forgetting of the first task when a new one is learned, we investigate elastic weight consolidation, a recently proposed method based on Fisher information, originally evaluated on reinforcement learning of Atari games. We use it to sequentially learn segmentation of normal brain structures and then segmentation of white matter lesions. Our findings show this recent method reduces catastrophic forgetting, while large room for improvement exists in these challenging settings for continual learning.

arXiv commentAccepted in Medical Imaging meets NIPS Workshop, NIPS 2018

References in corpus (2)

• Efficient Multi-Scale 3D CNN with Fully Connected CRF for Accurate Brain Lesion Segmentation
• Learning multiple visual domains with residual adapters

Cited by in corpus (10)

• A Survey on Active Learning and Human-in-the-Loop Deep Learning for Medical Image Analysis
• DoFE: Domain-oriented Feature Embedding for Generalizable Fundus Image Segmentation on Unseen Datasets
• Artificial Intelligence in Glioma Imaging: Challenges and Advances
• Online Coreset Selection for Rehearsal-based Continual Learning
• Domain-incremental Cardiac Image Segmentation with Style-oriented Replay and Domain-sensitive Feature Whitening
• Uncertainty Quantification in Deep Learning for Safer Neuroimage Enhancement
• Continual learning of longitudinal health records
• Exploring UMAP in hybrid models of entropy-based and representativeness sampling for active learning in biomedical segmentation
• Towards continuous learning for glioma segmentation with elastic weight consolidation
• Targeted Gradient Descent: A Novel Method for Convolutional Neural Networks Fine-tuning and Online-learning