Holistic decomposition convolution for effective semantic segmentation of medical volume images

Zeng, Guodong; Zheng, Guoyan (2019). Holistic decomposition convolution for effective semantic segmentation of medical volume images. Medical image analysis, 57, pp. 149-164. Elsevier 10.1016/j.media.2019.07.003

[img] Text
1-s2.0-S136184151930060X-main.pdf - Published Version
Restricted to registered users only
Available under License Publisher holds Copyright.

Download (5MB) | Request a copy

Convolutional Neural Networks (CNNs) have achieved state-of-the-art performance in many different 2D medical image analysis tasks. In clinical practice, however, a large part of the medical imaging data available is in 3D, e.g, magnetic resonance imaging (MRI) data, computed tomography (CT) data and data generated by many other modalities. This has motivated the development of 3D CNNs for volumetric image segmentation in order to benefit from more spatial context. Due to GPU memory restrictions caused by moving to fully 3D, state-of-the-art methods depend on subvolume/patch processing and the size of the input patch is usually small, limiting the incorporation of larger context information for a better performance. In this paper, we propose a novel Holistic Decomposition Convolution (HDC), which learns a number of separate kernels within the same layer and can be regarded as an inverse operation to the previously introduced Dense Upsampling Convolution (DUC), for an effective and efficient semantic segmentation of medical volume images. HDC consists of a periodic down-shuffling operation followed by a conventional 3D convolution. HDC has the advantage of significantly reducing the size of the data for sub-sequential processing while using all the information available in the input irrespective of the down-shuffling factors. We apply HDC directly to the input data, whose output will be used as the input to sub-sequential CNNs. In order to achieve volumetric dense prediction at final output, we need to recover full resolution, which is done by using DUC. We show that both HDC and DUC are network agnostic and can be combined with different CNNs for an improved performance in both training and testing phases. Results obtained from comprehensive experiments conducted on both MRI and CT data of different anatomical regions demonstrate the efficacy of the present approach.

Item Type:

Journal Article (Original Article)

Division/Institute:

04 Faculty of Medicine > Pre-clinic Human Medicine > Institute for Surgical Technology & Biomechanics ISTB
10 Strategic Research Centers > ARTORG Center for Biomedical Engineering Research
04 Faculty of Medicine > Faculty Institutions > sitem Center for Translational Medicine and Biomedical Entrepreneurship > Personalised Medicine

Graduate School:

Graduate School for Cellular and Biomedical Sciences (GCB)

UniBE Contributor:

Zeng, Guodong and Zheng, Guoyan

Subjects:

500 Science > 570 Life sciences; biology
600 Technology > 610 Medicine & health
000 Computer science, knowledge & systems

ISSN:

1361-8415

Publisher:

Elsevier

Language:

English

Submitter:

Nicolas Gerber

Date Deposited:

04 Sep 2019 10:38

Last Modified:

23 Oct 2019 12:30

Publisher DOI:

10.1016/j.media.2019.07.003

PubMed ID:

31302511

BORIS DOI:

10.7892/boris.132927

URI:

https://boris.unibe.ch/id/eprint/132927

Actions (login required)

Edit item Edit item
Provide Feedback