TBU Publications
Repository of TBU Publications

ISLES challenge: U-shaped convolution neural network with dilated convolution for 3D stroke lesion segmentation

DSpace Repository

Show simple item record


dc.title ISLES challenge: U-shaped convolution neural network with dilated convolution for 3D stroke lesion segmentation en
dc.contributor.author Turečková, Alžběta
dc.contributor.author Rodríguez-Sánchez, Antonio Jose
dc.relation.ispartof Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
dc.identifier.issn 0302-9743 Scopus Sources, Sherpa/RoMEO, JCR
dc.identifier.isbn 978-3-03-011722-1
dc.date.issued 2019
utb.relation.volume 11383 LNCS
dc.citation.spage 319
dc.citation.epage 327
dc.event.title 4th International MICCAI Brainlesion Workshop, BrainLes 2018 held in conjunction with the Medical Image Computing for Computer Assisted Intervention Conference, MICCAI 2018
dc.event.location Granada
utb.event.state-en Spain
utb.event.state-cs Španělsko
dc.event.sdate 2018-09-16
dc.event.edate 2018-09-20
dc.type conferenceObject
dc.language.iso en
dc.publisher Springer Verlag
dc.identifier.doi 10.1007/978-3-030-11723-8_32
dc.subject medical image segmentation en
dc.subject deep convolutional neural networks en
dc.subject U-Net en
dc.subject dilated convolution en
dc.description.abstract In this paper, we propose the algorithm for stroke lesion segmentation based on a deep convolutional neural network (CNN). The model is based on U-shaped CNN, which has been applied successfully to other medical image segmentation tasks. The network architecture was derived from the model presented in Isensee et al. [1] and is capable of processing whole 3D images. The model incorporates the convolution layers through upsampled filters – also known as dilated convolution. This change enlarges filter’s field of the view and allows the net to integrate larger context into the computation. We add the dilated convolution into different parts of network architecture and study the impact on the overall model performance. The best model which uses the dilated convolution in the input of the net outperforms the original architecture in nearly all used evaluation metrics. The code and trained models can be found on the GitHub website: http://github.com/tureckova/ISLES2018/. © Springer Nature Switzerland AG 2019. en
utb.faculty Faculty of Applied Informatics
dc.identifier.uri http://hdl.handle.net/10563/1008607
utb.identifier.obdid 43879270
utb.identifier.scopus 2-s2.0-85063451535
utb.identifier.wok 000612997600032
utb.source d-scopus
dc.date.accessioned 2019-07-08T11:59:58Z
dc.date.available 2019-07-08T11:59:58Z
dc.description.sponsorship Internal Grant Agency of Tomas Bata University [IGA/CebiaTech/2018/003]
utb.contributor.internalauthor Turečková, Alžběta
utb.fulltext.affiliation Alzbeta Tureckova 1, Antonio J. Rodríguez-Sánchez 2 1 Faculty of Applied Informatics, Tomas Bata University in Zlin, Nam T.G. Masaryka 5555, 760 01 Zlin, Czech Republic tureckova@utb.cz 2 Intelligent and Interactive Systems, Institute of Computer Science, University of Innsbruck, Innsbruck, Austria
utb.fulltext.dates -
utb.wos.affiliation [Tureckova, Alzbeta] Tomas Bata Univ Zlin, Fac Appl Informat, Nam TG Masaryka 5555, Zlin 76001, Czech Republic; [Rodriguez-Sanchez, Antonio J.] Univ Innsbruck, Inst Comp Sci, Intelligent & Interact Syst, Innsbruck, Austria
utb.scopus.affiliation Faculty of Applied Informatics, Tomas Bata University in Zlin, Nam T.G. Masaryka 5555, Zlin, 760 01, Czech Republic; Intelligent and Interactive Systems, Institute of Computer Science, University of Innsbruck, Innsbruck, Austria
utb.fulltext.faculty Faculty of Applied Informatics
Find Full text

Files in this item

Show simple item record