Automated image curation in diabetic retinopathy screening using deep learning.
Journal
Scientific reports
ISSN: 2045-2322
Titre abrégé: Sci Rep
Pays: England
ID NLM: 101563288
Informations de publication
Date de publication:
01 07 2022
01 07 2022
Historique:
received:
08
02
2022
accepted:
24
06
2022
entrez:
1
7
2022
pubmed:
2
7
2022
medline:
7
7
2022
Statut:
epublish
Résumé
Diabetic retinopathy (DR) screening images are heterogeneous and contain undesirable non-retinal, incorrect field and ungradable samples which require curation, a laborious task to perform manually. We developed and validated single and multi-output laterality, retinal presence, retinal field and gradability classification deep learning (DL) models for automated curation. The internal dataset comprised of 7743 images from DR screening (UK) with 1479 external test images (Portugal and Paraguay). Internal vs external multi-output laterality AUROC were right (0.994 vs 0.905), left (0.994 vs 0.911) and unidentifiable (0.996 vs 0.680). Retinal presence AUROC were (1.000 vs 1.000). Retinal field AUROC were macula (0.994 vs 0.955), nasal (0.995 vs 0.962) and other retinal field (0.997 vs 0.944). Gradability AUROC were (0.985 vs 0.918). DL effectively detects laterality, retinal presence, retinal field and gradability of DR screening images with generalisation between centres and populations. DL models could be used for automated image curation within DR screening.
Identifiants
pubmed: 35778615
doi: 10.1038/s41598-022-15491-1
pii: 10.1038/s41598-022-15491-1
pmc: PMC9249740
doi:
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
11196Subventions
Organisme : Diabetes UK
ID : 20/0006144
Pays : United Kingdom
Informations de copyright
© 2022. The Author(s).
Références
Ophthalmology. 2018 Aug;125(8):1264-1272
pubmed: 29548646
JAMA. 2016 Dec 13;316(22):2402-2410
pubmed: 27898976
Data Brief. 2021 Apr 21;36:107068
pubmed: 34307801
Nat Biomed Eng. 2022 Mar 29;:
pubmed: 35352000
Annu Int Conf IEEE Eng Med Biol Soc. 2011;2011:5959-64
pubmed: 22255697
Ophthalmology. 1991 May;98(5 Suppl):766-85
pubmed: 2062512
Sci Rep. 2021 Jan 12;11(1):586
pubmed: 33436781
Lancet Digit Health. 2021 Jan;3(1):e51-e66
pubmed: 33735069
Ophthalmic Res. 2018;60(1):9-17
pubmed: 29339646
PLoS One. 2019 Sep 19;14(9):e0222025
pubmed: 31536537
Clin Ophthalmol. 2020 Feb 13;14:419-429
pubmed: 32103888
Ophthalmol Retina. 2020 Aug;4(8):793-800
pubmed: 32362553
J Biomed Opt. 2012 Jul;17(7):076021
pubmed: 22894504
Eye Vis (Lond). 2015 Sep 30;2:17
pubmed: 26605370
Diabetes Res Clin Pract. 2019 Nov;157:107843
pubmed: 31518657
J Digit Imaging. 2018 Dec;31(6):923-928
pubmed: 29948436
NPJ Digit Med. 2019 Apr 10;2:25
pubmed: 31304372
J Med Imaging (Bellingham). 2014 Apr;1(1):014001
pubmed: 26158021
J Biomed Opt. 2014 Apr;19(4):046006
pubmed: 24718384
Sci Rep. 2021 May 4;11(1):9469
pubmed: 33947946
Clin Exp Ophthalmol. 2016 May;44(4):260-77
pubmed: 26716602
Transl Vis Sci Technol. 2021 Sep 1;10(11):16
pubmed: 34524409
Acta Diabetol. 2017 Jun;54(6):515-525
pubmed: 28224275
Conf Proc IEEE Eng Med Biol Soc. 2005;2006:162-5
pubmed: 17282136
Int J Comput Assist Radiol Surg. 2010 Nov;5(6):557-64
pubmed: 20490705
Nat Commun. 2021 May 28;12(1):3242
pubmed: 34050158
Comput Biol Med. 2016 Apr 1;71:67-76
pubmed: 26894596
Invest Ophthalmol Vis Sci. 2016 Oct 01;57(13):5200-5206
pubmed: 27701631
Acta Ophthalmol. 2009 Sep;87(6):643-7
pubmed: 19719806
Lancet Glob Health. 2021 Feb;9(2):e144-e160
pubmed: 33275949
Biometrics. 1988 Sep;44(3):837-45
pubmed: 3203132
Invest Ophthalmol Vis Sci. 2006 Mar;47(3):1120-5
pubmed: 16505050
Acta Ophthalmol. 2018 Feb;96(1):63-68
pubmed: 29178249
BMJ Open. 2017 Jul 18;7(7):e015024
pubmed: 28720613
Comput Biol Med. 2019 May;108:317-331
pubmed: 31028967
Comput Biol Med. 2018 Dec 1;103:64-70
pubmed: 30340214