Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/67199
Title: 3D FractalNet : dense volumetric segmentation for cardiovascular MRI volumes
Authors: Yu, L
Yang, X
Qin, J 
Heng, PA
Issue Date: 2017
Publisher: Springer
Source: Lecture notes in computer science (including subseries Lecture notes in artificial intelligence and lecture notes in bioinformatics), 2017, v. 10129, p. 103-110 How to cite?
Journal: Lecture notes in computer science (including subseries Lecture notes in artificial intelligence and lecture notes in bioinformatics) 
Abstract: Cardiac image segmentation plays a crucial role in various medical applications. However, differentiating branchy structures and slicing fuzzy boundaries from cardiovascular MRI volumes remain very challenging tasks. In this paper, we propose a novel deeply-supervised 3D fractal network for efficient automated whole heart and great vessel segmentation in MRI volumes. The proposed 3D fractal network takes advantage of fully convolutional architecture to perform efficient, precise and volume-to-volume prediction. Notably, by recursively applying a single expansion rule, we construct our network in a novel self-similar fractal scheme and thus promote it in combining hierarchical clues for accurate segmentation. More importantly, we employ deep supervision mechanism to alleviate the vanishing gradients problem and improve the training efficiency of our network on small medical image dataset. We evaluated our method on the HVSMR 2016 Challenge dataset. Extensive experimental results demonstrated the superior performance of our method, ranking top in both two phases.
Description: 1st International Workshops on Reconstruction and Analysis of Moving Body Organs, RAMBO 2016 and 1st International Workshops on Whole-Heart and Great Vessel Segmentation from 3D Cardiovascular MRI in Congenital Heart Disease, HVSMR 2016 Held in Conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016, Greece, 17-21 October, 2016
URI: http://hdl.handle.net/10397/67199
ISBN: 9783319522791
ISSN: 0302-9743
EISSN: 1611-3349
DOI: 10.1007/978-3-319-52280-7_10
Appears in Collections:Conference Paper

Access
View full-text via PolyU eLinks SFX Query
Show full item record

SCOPUSTM   
Citations

1
Last Week
0
Last month
Citations as of Oct 10, 2017

Page view(s)

159
Last Week
4
Last month
Checked on Oct 15, 2017

Google ScholarTM

Check

Altmetric



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