[139390]
Title: 4D Deep Learning for Multiple-Sclerosis Lesion Activity Segmentation Medical Imaging with Deep Learning
Written by: N. Gessert and M. Bengs and J. Kr├╝ger and R. Opfer and A.-C. Ostwaldt and P. Manogaran and S. Schippling and A. Schlaefer
in: 2020
Volume: Number:
on pages: accepted
Chapter:
Editor:
Publisher:
Series:
Address:
Edition:
ISBN:
how published:
Organization:
School:
Institution:
Type:
DOI:
URL: https://openreview.net/forum?id=sMsAIWBSvg
ARXIVID:
PMID:

[www] [BibTex]

Note:

Abstract: Multiple sclerosis lesion activity segmentation is the task of detecting new and enlarging lesions that appeared between a baseline and a follow\-up brain MRI scan. While deep learning methods for single\-scan lesion segmentation are common, deep learning approaches for lesion activity have only been proposed recently. Here, a two\-path architecture processes two 3D MRI volumes from two time points. In this work, we investigate whether extending this problem to full 4D deep learning using a history of MRI volumes and thus an extended baseline can improve performance. For this purpose, we design a recurrent multi\-encoder\-decoder architecture for processing 4D data. We find that adding more temporal information is beneficial and our proposed architecture outperforms previous approaches with a lesion\-wise true positive rate of 0.84 at a lesion\-wise false positive rate of 0.19.

To top