Optimized High Resolution 3D Dense-U-Net Network for Brain and Spine Segmentation

Simple item page
dc.contributor.authorKolařík, Martincs
dc.contributor.authorBurget, Radimcs
dc.contributor.authorUher, Václavcs
dc.contributor.authorŘíha, Kamilcs
dc.contributor.authorDutta, Malay Kishorecs
dc.coverage.issue3cs
dc.coverage.volume9cs
dc.date.issued2019-02-15cs
dc.description.abstractThe 3D image segmentation is the process of partitioning a digital 3D volumes into multiple segments. This paper presents a fully automatic method for high resolution 3D volumetric segmentation of medical image data using modern supervised deep learning approach. We introduce 3D Dense-U-Net neural network architecture implementing densely connected layers. It has been optimized for graphic process unit accelerated high resolution image processing on currently available hardware (Nvidia GTX 1080ti). The method has been evaluated on MRI brain 3D volumetric dataset and CT thoracic scan dataset for spine segmentation. In contrast with many previous methods, our approach is capable of precise segmentation of the input image data in the original resolution, without any pre-processing of the input image. It can process image data in 3D and has achieved accuracy of 99.72% on MRI brain dataset, which outperformed results achieved by human expert. On lumbar and thoracic vertebrae CT dataset it has achieved the accuracy of 99.80%. The architecture proposed in this paper can also be easily applied to any task already using U-Net network as a segmentation algorithm to enhance its results. Complete source code was released online under open-source license.en
dc.formattextcs
dc.format.extent1-17cs
dc.format.mimetypeapplication/pdfcs
dc.identifier.citationApplied Sciences - Basel. 2019, vol. 9, issue 3, p. 1-17.en
dc.identifier.doi10.3390/app9030404cs
dc.identifier.issn2076-3417cs
dc.identifier.orcid0000-0001-6158-6162cs
dc.identifier.orcid0000-0003-1849-5390cs
dc.identifier.orcid0000-0001-7453-7941cs
dc.identifier.orcid0000-0002-6196-5215cs
dc.identifier.other155280cs
dc.identifier.researcheridB-9326-2019cs
dc.identifier.scopus23011250200cs
dc.identifier.urihttp://hdl.handle.net/11012/179271
dc.language.isoencs
dc.publisherMDPIcs
dc.relation.ispartofApplied Sciences - Baselcs
dc.relation.urihttps://www.mdpi.com/2076-3417/9/3/404cs
dc.rightsCreative Commons Attribution 4.0 Internationalcs
dc.rights.accessopenAccesscs
dc.rights.sherpahttp://www.sherpa.ac.uk/romeo/issn/2076-3417/cs
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/cs
dc.subject3D segmentationen
dc.subjectbrainen
dc.subjectdeep learningen
dc.subjectneural networken
dc.subjectopen-sourceen
dc.subjectsemantic segmentationen
dc.subjectspineen
dc.subjectu-neten
dc.titleOptimized High Resolution 3D Dense-U-Net Network for Brain and Spine Segmentationen
dc.type.driverarticleen
dc.type.statusPeer-revieweden
dc.type.versionpublishedVersionen
sync.item.dbidVAV-155280en
sync.item.dbtypeVAVen
sync.item.insts2025.02.03 15:39:19en
sync.item.modts2025.01.17 16:39:26en
thesis.grantorVysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií. Ústav telekomunikacícs
thesis.grantorVysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií. oddělení-TKO-SIXcs
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
applsci0900404v2.pdf
Size:
3.2 MB
Format:
Adobe Portable Document Format
Description:
applsci0900404v2.pdf
Download
Collections
Ústav telekomunikací
oddělení-TKO-SIX