King's College London

Research portal

On the compactness, efficiency, and representation of 3D convolutional networks: Brain parcellation as a pretext task

Research output: Chapter in Book/Report/Conference proceedingConference paper

Original languageEnglish
Title of host publicationInformation Processing in Medical Imaging - 25th International Conference, IPMI 2017, Proceedings
PublisherSpringer Verlag
Pages348-360
Number of pages13
Volume10265 LNCS
ISBN (Print)9783319590493
DOIs
Publication statusPublished - 1 Jan 2017
Event25th International Conference on Information Processing in Medical Imaging, IPMI 2017 - Boone, United States
Duration: 25 Jun 201730 Jun 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10265 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference25th International Conference on Information Processing in Medical Imaging, IPMI 2017
CountryUnited States
CityBoone
Period25/06/201730/06/2017

King's Authors

Abstract

Deep convolutional neural networks are powerful tools for learning visual representations from images. However, designing efficient deep architectures to analyse volumetric medical images remains challenging. This work investigates efficient and flexible elements of modern convolutional networks such as dilated convolution and residual connection. With these essential building blocks, we propose a high-resolution, compact convolutional network for volumetric image segmentation. To illustrate its efficiency of learning 3D representation from large-scale image data, the proposed network is validated with the challenging task of parcellating 155 neuroanatomical structures from brain MR images. Our experiments show that the proposed network architecture compares favourably with state-of-the-art volumetric segmentation networks while being an order of magnitude more compact. We consider the brain parcellation task as a pretext task for volumetric image segmentation; our trained network potentially provides a good starting point for transfer learning. Additionally, we show the feasibility of voxel-level uncertainty estimation using a sampling approximation through dropout.

View graph of relations

© 2018 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454