A CNN-aided method to predict glaucoma progression using DARC (Detection of Apoptosing Retinal Cells).
Adult
Aged
Algorithms
Annexin A5
/ administration & dosage
Apoptosis
Automation
Clinical Trials, Phase II as Topic
Disease Progression
Female
Glaucoma
/ pathology
Humans
Image Processing, Computer-Assisted
Male
Middle Aged
Neural Networks, Computer
Observer Variation
Retinal Ganglion Cells
/ pathology
Tomography, Optical Coherence
Artificial Intelligence
Biomarker
CNN
apoptosis
glaucoma
imaging
Journal
Expert review of molecular diagnostics
ISSN: 1744-8352
Titre abrégé: Expert Rev Mol Diagn
Pays: England
ID NLM: 101120777
Informations de publication
Date de publication:
07 2020
07 2020
Historique:
pubmed:
21
4
2020
medline:
22
9
2021
entrez:
21
4
2020
Statut:
ppublish
Résumé
A key objective in glaucoma is to identify those at risk of rapid progression and blindness. Recently, a novel first-in-man method for visualising apoptotic retinal cells called DARC (Detection-of-Apoptosing-Retinal-Cells) was reported. The aim was to develop an automatic CNN-aided method of DARC spot detection to enable prediction of glaucoma progression. Anonymised DARC images were acquired from healthy control (n=40) and glaucoma (n=20) Phase 2 clinical trial subjects (ISRCTN10751859) from which 5 observers manually counted spots. The CNN-aided algorithm was trained and validated using manual counts from control subjects, and then tested on glaucoma eyes. The algorithm had 97.0% accuracy, 91.1% sensitivity and 97.1% specificity to spot detection when compared to manual grading of 50% controls. It was next tested on glaucoma patient eyes defined as progressing or stable based on a significant (p<0.05) rate of progression using OCT-retinal nerve fibre layer measurements at 18 months. It demonstrated 85.7% sensitivity, 91.7% specificity with AUC of 0.89, and a significantly (p=0.0044) greater DARC count in those patients who later progressed. This CNN-enabled algorithm provides an automated and objective measure of DARC, promoting its use as an AI-aided biomarker for predicting glaucoma progression and testing new drugs.
Sections du résumé
BACKGROUND
A key objective in glaucoma is to identify those at risk of rapid progression and blindness. Recently, a novel first-in-man method for visualising apoptotic retinal cells called DARC (Detection-of-Apoptosing-Retinal-Cells) was reported. The aim was to develop an automatic CNN-aided method of DARC spot detection to enable prediction of glaucoma progression.
METHODS
Anonymised DARC images were acquired from healthy control (n=40) and glaucoma (n=20) Phase 2 clinical trial subjects (ISRCTN10751859) from which 5 observers manually counted spots. The CNN-aided algorithm was trained and validated using manual counts from control subjects, and then tested on glaucoma eyes.
RESULTS
The algorithm had 97.0% accuracy, 91.1% sensitivity and 97.1% specificity to spot detection when compared to manual grading of 50% controls. It was next tested on glaucoma patient eyes defined as progressing or stable based on a significant (p<0.05) rate of progression using OCT-retinal nerve fibre layer measurements at 18 months. It demonstrated 85.7% sensitivity, 91.7% specificity with AUC of 0.89, and a significantly (p=0.0044) greater DARC count in those patients who later progressed.
CONCLUSION
This CNN-enabled algorithm provides an automated and objective measure of DARC, promoting its use as an AI-aided biomarker for predicting glaucoma progression and testing new drugs.
Identifiants
pubmed: 32310684
doi: 10.1080/14737159.2020.1758067
pmc: PMC7115906
mid: EMS86250
doi:
Substances chimiques
Annexin A5
0
Types de publication
Comparative Study
Journal Article
Research Support, Non-U.S. Gov't
Validation Study
Langues
eng
Sous-ensembles de citation
IM
Pagination
737-748Subventions
Organisme : Wellcome Trust
ID : 088029
Pays : United Kingdom
Organisme : Wellcome Trust
ID : 099729
Pays : United Kingdom
Références
Invest Ophthalmol Vis Sci. 2007 Mar;48(3):1156-63
pubmed: 17325159
Bioinformatics. 2020 Mar 1;36(5):1542-1552
pubmed: 31591638
Br J Ophthalmol. 2003 Jun;87(6):726-30
pubmed: 12770970
Brain. 2017 Jun 01;140(6):1757-1767
pubmed: 28449038
Lancet. 2017 Nov 11;390(10108):2183-2193
pubmed: 28577860
Invest Ophthalmol Vis Sci. 2015 Jan 27;56(2):1253-60
pubmed: 25626965
Ophthalmology. 2017 Dec;124(12S):S57-S65
pubmed: 29157363
Am J Ophthalmol. 2019 Feb;198:136-145
pubmed: 30316669
J Digit Imaging. 2019 Oct;32(5):888-896
pubmed: 30838482
Annu Int Conf IEEE Eng Med Biol Soc. 2018 Jul;2018:862-865
pubmed: 30440527
Nat Biomed Eng. 2018 Mar;2(3):158-164
pubmed: 31015713
Ophthalmology. 2006 Dec;113(12):2137-43
pubmed: 16996592
PLoS One. 2019 Jan 10;14(1):e0209409
pubmed: 30629635
Ophthalmology. 2019 Apr;126(4):513-521
pubmed: 30578810
Ophthalmology. 2012 Apr;119(4):731-7
pubmed: 22264886
Ophthalmology. 2016 Jun;123(6):1173-80
pubmed: 26949119
Ophthalmic Physiol Opt. 2013 Jan;33(1):7-14
pubmed: 23252852
Am J Ophthalmol. 2020 May;213:203-216
pubmed: 31899204
Ophthalmology. 2009 Dec;116(12):2294-9
pubmed: 19800694
Am J Ophthalmol. 2007 Aug;144(2):266-275
pubmed: 17543874
Invest Ophthalmol Vis Sci. 2006 Sep;47(9):3896-903
pubmed: 16936102
Invest Ophthalmol Vis Sci. 2019 Jan 2;60(1):365-375
pubmed: 30682206
Invest Ophthalmol Vis Sci. 2011 Oct 04;52(11):7842-51
pubmed: 21972262
Med Image Anal. 2017 Jul;39:178-193
pubmed: 28511066
Invest Ophthalmol Vis Sci. 2018 Jul 2;59(8):3199-3208
pubmed: 29971444
Am J Ophthalmol. 2019 May;201:9-18
pubmed: 30689990
Ophthalmology. 2018 Dec;125(12):1874-1885
pubmed: 30082073
Invest Ophthalmol Vis Sci. 2009 Apr;50(4):1497-505
pubmed: 19321793
Invest Ophthalmol Vis Sci. 2007 Feb;48(2):763-73
pubmed: 17251476
Arch Ophthalmol. 2002 Jun;120(6):714-20; discussion 829-30
pubmed: 12049575
Ophthalmology. 2015 Sep;122(9):1786-94
pubmed: 26198806
Nature. 2020 Jan;577(7788):89-94
pubmed: 31894144
Br J Ophthalmol. 2006 Mar;90(3):262-7
pubmed: 16488940
Transl Vis Sci Technol. 2016 Nov 01;5(6):1
pubmed: 27847688
Ophthalmology. 2014 Nov;121(11):2081-90
pubmed: 24974815
Korean J Ophthalmol. 2015 Feb;29(1):40-6
pubmed: 25646059
Ophthalmol Retina. 2018 Jan;2(1):24-30
pubmed: 31047298
Invest Ophthalmol Vis Sci. 2013 Aug 15;54(8):5505-13
pubmed: 23833069
Br J Ophthalmol. 2011 Sep;95(9):1276-83
pubmed: 21131377
PLoS One. 2019 Sep 12;14(9):e0221720
pubmed: 31513631
J Glaucoma. 2012 Apr-May;21(4):266-73
pubmed: 21637116
PLoS One. 2018 Nov 28;13(11):e0206081
pubmed: 30485270
Am J Ophthalmol. 2017 Sep;181:106-113
pubmed: 28669780
JAMA Ophthalmol. 2017 Apr 01;135(4):313-319
pubmed: 28241171
Invest Ophthalmol Vis Sci. 2016 Jul 1;57(9):OCT80-5
pubmed: 27409509
J Glaucoma. 2013 Oct-Nov;22(8):652-8
pubmed: 22576170
Br J Ophthalmol. 2019 Feb;103(2):167-175
pubmed: 30361278
Am J Ophthalmol. 2017 Aug;180:151-157
pubmed: 28624324