Self-supervised pretraining for transferable quantitative phase image cell segmentation

dc.contributor.authorVičar, Tomášcs
dc.contributor.authorChmelík, Jiřícs
dc.contributor.authorJakubíček, Romancs
dc.contributor.authorChmelíková, Larisacs
dc.contributor.authorGumulec, Jaromírcs
dc.contributor.authorBalvan, Jancs
dc.contributor.authorProvazník, Valentýnacs
dc.contributor.authorKolář, Radimcs
dc.coverage.issue10cs
dc.coverage.volume12cs
dc.date.issued2021-09-24cs
dc.description.abstractIn this paper, a novel U-Net-based method for robust adherent cell segmentation for quantitative phase microscopy image is designed and optimised. We designed and evaluated four specific post-processing pipelines. To increase the transferability to different cell types, non-deep learning transfer with adjustable parameters is used in the post-processing step. Additionally, we proposed a self-supervised pretraining technique using nonlabelled data, which is trained to reconstruct multiple image distortions and improved the segmentation performance from 0.67 to 0.70 of object-wise intersection over union. Moreover, we publish a new dataset of manually labelled images suitable for this task together with the unlabelled data for self-supervised pretraining.en
dc.formattextcs
dc.format.extent6514-6528cs
dc.format.mimetypeapplication/pdfcs
dc.identifier.citationBiomedical Optics Express. 2021, vol. 12, issue 10, p. 6514-6528.en
dc.identifier.doi10.1364/BOE.433212cs
dc.identifier.issn2156-7085cs
dc.identifier.orcid0000-0002-9136-7873cs
dc.identifier.orcid0000-0001-9950-6279cs
dc.identifier.orcid0000-0003-4293-260Xcs
dc.identifier.orcid0000-0002-3178-4202cs
dc.identifier.orcid0000-0002-9658-3444cs
dc.identifier.orcid0000-0002-3422-7938cs
dc.identifier.orcid0000-0002-0469-6397cs
dc.identifier.other172596cs
dc.identifier.researcheridC-6006-2018cs
dc.identifier.researcheridH-9359-2017cs
dc.identifier.researcheridD-3622-2018cs
dc.identifier.researcheridD-3886-2018cs
dc.identifier.researcheridD-7638-2012cs
dc.identifier.researcheridF-4121-2012cs
dc.identifier.researcheridC-8547-2014cs
dc.identifier.scopus57202426072cs
dc.identifier.scopus57188877911cs
dc.identifier.scopus57188881119cs
dc.identifier.scopus55769747816cs
dc.identifier.scopus6701729526cs
dc.identifier.urihttp://hdl.handle.net/11012/201741
dc.language.isoencs
dc.publisherOptica Publishing Groupcs
dc.relation.ispartofBiomedical Optics Expresscs
dc.relation.urihttps://www.osapublishing.org/boe/fulltext.cfm?uri=boe-12-10-6514&id=459853cs
dc.rights(C) Optica Publishing Groupcs
dc.rights.accessopenAccesscs
dc.rights.sherpahttp://www.sherpa.ac.uk/romeo/issn/2156-7085/cs
dc.subjectcell segmentationen
dc.subjectdeep learningen
dc.subjecttransfer learningen
dc.subjectself-superviseden
dc.titleSelf-supervised pretraining for transferable quantitative phase image cell segmentationen
dc.type.driverarticleen
dc.type.statusPeer-revieweden
dc.type.versionpublishedVersionen
sync.item.dbidVAV-172596en
sync.item.dbtypeVAVen
sync.item.insts2025.02.03 15:39:50en
sync.item.modts2025.01.17 15:15:40en
thesis.grantorVysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií. Ústav biomedicínského inženýrstvícs
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
boe12106514.pdf
Size:
7.83 MB
Format:
Adobe Portable Document Format
Description:
boe12106514.pdf