Towards continual learning in medical imaging
File(s)
Author(s)
Baweja, C
Glocker, B
Kamnitsas, K
Type
Working Paper
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.
Online Publication Date
2019-07-10T13:54:53Z
Source Database
arxiv
Identifier
http://arxiv.org/abs/1811.02496v1
Subjects
cs.CV
cs.CV
cs.LG
Notes
Accepted in Medical Imaging meets NIPS Workshop, NIPS 2018