Self-supervised pretraining for transferable quantitative phase image cell segmentation.


Journal

Biomedical optics express
ISSN: 2156-7085
Titre abrégé: Biomed Opt Express
Pays: United States
ID NLM: 101540630

Informations de publication

Date de publication:
01 Oct 2021
Historique:
received: 04 06 2021
revised: 03 08 2021
accepted: 24 08 2021
entrez: 8 11 2021
pubmed: 9 11 2021
medline: 9 11 2021
Statut: epublish

Résumé

In 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.

Identifiants

pubmed: 34745753
doi: 10.1364/BOE.433212
pii: 433212
pmc: PMC8547997
doi:

Types de publication

Journal Article

Langues

eng

Pagination

6514-6528

Informations de copyright

© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement.

Déclaration de conflit d'intérêts

The authors declare no conflicts of interest.

Références

J Biomed Opt. 2020 Nov;25(11):
pubmed: 33188571
Nat Commun. 2016 Jan 07;7:10256
pubmed: 26739115
Opt Express. 2013 Jun 17;21(12):14747-62
pubmed: 23787662
Biomed Opt Express. 2021 Mar 01;12(3):1683-1706
pubmed: 33796381
Cytometry A. 2017 May;91(5):470-481
pubmed: 28264140
Nat Methods. 2017 Dec;14(12):1141-1152
pubmed: 29083403
Sci Rep. 2020 Jan 31;10(1):1566
pubmed: 32005874
IEEE Trans Med Imaging. 2018 Apr;37(4):929-940
pubmed: 29610072
Biomed Opt Express. 2017 Sep 12;8(10):4466-4479
pubmed: 29082078
Cytometry A. 2017 May;91(5):482-493
pubmed: 28426133
Cytometry A. 2019 Sep;95(9):952-965
pubmed: 31313519
IEEE J Biomed Health Inform. 2013 May;17(3):642-53
pubmed: 24592465
Cytometry A. 2017 May;91(5):443-449
pubmed: 28371011
Biomed Opt Express. 2021 Mar 31;12(4):2519-2530
pubmed: 33996245
PLoS One. 2020 Dec 8;15(12):e0243219
pubmed: 33290432
BMC Bioinformatics. 2019 Jun 28;20(1):360
pubmed: 31253078
Nat Methods. 2021 Jan;18(1):100-106
pubmed: 33318659

Auteurs

Tomas Vicar (T)

Department of Biomedical Engineering, Faculty of Electrical Engineering and Communication, Brno University of Technology, Brno, Czech Republic.
Department of Pathological Physiology, Faculty of Medicine, Masaryk University, Brno, Czech Republic.

Jiri Chmelik (J)

Department of Biomedical Engineering, Faculty of Electrical Engineering and Communication, Brno University of Technology, Brno, Czech Republic.

Roman Jakubicek (R)

Department of Biomedical Engineering, Faculty of Electrical Engineering and Communication, Brno University of Technology, Brno, Czech Republic.

Larisa Chmelikova (L)

Department of Biomedical Engineering, Faculty of Electrical Engineering and Communication, Brno University of Technology, Brno, Czech Republic.

Jaromir Gumulec (J)

Department of Pathological Physiology, Faculty of Medicine, Masaryk University, Brno, Czech Republic.

Jan Balvan (J)

Department of Pathological Physiology, Faculty of Medicine, Masaryk University, Brno, Czech Republic.

Ivo Provaznik (I)

Department of Biomedical Engineering, Faculty of Electrical Engineering and Communication, Brno University of Technology, Brno, Czech Republic.

Radim Kolar (R)

Department of Biomedical Engineering, Faculty of Electrical Engineering and Communication, Brno University of Technology, Brno, Czech Republic.

Classifications MeSH