Multi-modal learning from unpaired images: Application to multi-organ segmentation in CT and MRI

File Description SizeFormat 
multimodal.pdfAccepted version1.73 MBAdobe PDFView/Open
Title: Multi-modal learning from unpaired images: Application to multi-organ segmentation in CT and MRI
Authors: Valindria, V
Pawlowski, N
Rajchl, M
Lavdas, I
Aboagye, EO
Rockall, A
Rueckert, D
Glocker, B
Item Type: Conference Paper
Abstract: Convolutional neural networks have been widely used in medical image segmentation. The amount of training data strongly determines the overall performance. Most approaches are applied for a single imaging modality, e.g., brain MRI. In practice, it is often difficult to acquire sufficient training data of a certain imaging modality. The same anatomical structures, however, may be visible in different modalities such as major organs on abdominal CT and MRI. In this work, we investigate the effectiveness of learning from multiple modalities to improve the segmentation accuracy on each individual modality. We study the feasibility of using a dual-stream encoder-decoder architecture to learn modality-independent, and thus, generalisable and robust features. All of our MRI and CT data are unpaired, which means they are obtained from different subjects and not registered to each other. Experiments show that multi-modal learning can improve overall accuracy over modality-specific training. Results demonstrate that information across modalities can in particular improve performance on varying structures such as the spleen.
Issue Date: 7-May-2018
Date of Acceptance: 20-Jan-2018
URI: http://hdl.handle.net/10044/1/56452
DOI: https://doi.org/10.1109/WACV.2018.00066
ISBN: 9781538648872
Publisher: IEEE
Journal / Book Title: 2018 IEEE Winter Conference on Applications of Computer Vision (WACV)
Copyright Statement: © 2018 IEEE.
Sponsor/Funder: National Institute for Health Research
Microsoft Reseach
NVIDIA Corporation
Imperial College London
Imperial College Healthcare NHS Trust- BRC Funding
Imperial College Healthcare NHS Trust- BRC Funding
Funder's Grant Number: EME/13/122/01
Imperial College Research Fellowship
RDC04 79560
RDC04
Conference Name: IEEE Winter Conference on Applications of Computer Vision
Keywords: Science & Technology
Technology
Computer Science, Artificial Intelligence
Engineering, Electrical & Electronic
Computer Science
Engineering
NEURAL-NETWORKS
Publication Status: Published
Start Date: 2018-03-12
Finish Date: 2018-03-14
Conference Place: Lake Tahoe, California, USA
Online Publication Date: 2018-05-07
Appears in Collections:Faculty of Engineering
Division of Surgery
Computing
Division of Cancer
Faculty of Medicine



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Creative Commons