Tags

Type your tag names separated by a space and hit enter

Prospective evaluation of an artificial intelligence-enabled algorithm for automated diabetic retinopathy screening of 30 000 patients.
Br J Ophthalmol. 2020 Jun 30 [Online ahead of print]BJ

Abstract

BACKGROUND/AIMS

Human grading of digital images from diabetic retinopathy (DR) screening programmes represents a significant challenge, due to the increasing prevalence of diabetes. We evaluate the performance of an automated artificial intelligence (AI) algorithm to triage retinal images from the English Diabetic Eye Screening Programme (DESP) into test-positive/technical failure versus test-negative, using human grading following a standard national protocol as the reference standard.

METHODS

Retinal images from 30 405 consecutive screening episodes from three English DESPs were manually graded following a standard national protocol and by an automated process with machine learning enabled software, EyeArt v2.1. Screening performance (sensitivity, specificity) and diagnostic accuracy (95% CIs) were determined using human grades as the reference standard.

RESULTS

Sensitivity (95% CIs) of EyeArt was 95.7% (94.8% to 96.5%) for referable retinopathy (human graded ungradable, referable maculopathy, moderate-to-severe non-proliferative or proliferative). This comprises sensitivities of 98.3% (97.3% to 98.9%) for mild-to-moderate non-proliferative retinopathy with referable maculopathy, 100% (98.7%,100%) for moderate-to-severe non-proliferative retinopathy and 100% (97.9%,100%) for proliferative disease. EyeArt agreed with the human grade of no retinopathy (specificity) in 68% (67% to 69%), with a specificity of 54.0% (53.4% to 54.5%) when combined with non-referable retinopathy.

CONCLUSION

The algorithm demonstrated safe levels of sensitivity for high-risk retinopathy in a real-world screening service, with specificity that could halve the workload for human graders. AI machine learning and deep learning algorithms such as this can provide clinically equivalent, rapid detection of retinopathy, particularly in settings where a trained workforce is unavailable or where large-scale and rapid results are needed.

Authors+Show Affiliations

Moorfields Biomedical Research Centre, Moorfields Eye Hospital, London, UK.Moorfields Biomedical Research Centre, Moorfields Eye Hospital, London, UK. Institute of Ophthalmology, UCL, London, UK.Homerton University Hospital NHS Trust, London, UK.Homerton University Hospital NHS Trust, London, UK.Homerton University Hospital NHS Trust, London, UK.Gloucestershire Hospitals NHS Foundation Trust, Cheltenham, UK.Gloucestershire Hospitals NHS Foundation Trust, Cheltenham, UK.Gloucestershire Hospitals NHS Foundation Trust, Cheltenham, UK.Guy's and Saint Thomas' NHS Foundation Trust, London, UK.Guy's and Saint Thomas' NHS Foundation Trust, London, UK.Guy's and Saint Thomas' NHS Foundation Trust, London, UK.Population Health Research Institute, St George's, University of London, London, UK.Moorfields Biomedical Research Centre, Moorfields Eye Hospital, London, UK. Institute of Ophthalmology, UCL, London, UK.Population Health Research Institute, St George's, University of London, London, UK arudnick@sgul.ac.uk.

Pub Type(s)

Journal Article

Language

eng

PubMed ID

32606081

Citation

Heydon, Peter, et al. "Prospective Evaluation of an Artificial Intelligence-enabled Algorithm for Automated Diabetic Retinopathy Screening of 30 000 Patients." The British Journal of Ophthalmology, 2020.
Heydon P, Egan C, Bolter L, et al. Prospective evaluation of an artificial intelligence-enabled algorithm for automated diabetic retinopathy screening of 30 000 patients. Br J Ophthalmol. 2020.
Heydon, P., Egan, C., Bolter, L., Chambers, R., Anderson, J., Aldington, S., Stratton, I. M., Scanlon, P. H., Webster, L., Mann, S., du Chemin, A., Owen, C. G., Tufail, A., & Rudnicka, A. R. (2020). Prospective evaluation of an artificial intelligence-enabled algorithm for automated diabetic retinopathy screening of 30 000 patients. The British Journal of Ophthalmology. https://doi.org/10.1136/bjophthalmol-2020-316594
Heydon P, et al. Prospective Evaluation of an Artificial Intelligence-enabled Algorithm for Automated Diabetic Retinopathy Screening of 30 000 Patients. Br J Ophthalmol. 2020 Jun 30; PubMed PMID: 32606081.
* Article titles in AMA citation format should be in sentence-case
TY - JOUR T1 - Prospective evaluation of an artificial intelligence-enabled algorithm for automated diabetic retinopathy screening of 30 000 patients. AU - Heydon,Peter, AU - Egan,Catherine, AU - Bolter,Louis, AU - Chambers,Ryan, AU - Anderson,John, AU - Aldington,Steve, AU - Stratton,Irene M, AU - Scanlon,Peter Henry, AU - Webster,Laura, AU - Mann,Samantha, AU - du Chemin,Alain, AU - Owen,Christopher G, AU - Tufail,Adnan, AU - Rudnicka,Alicja Regina, Y1 - 2020/06/30/ PY - 2020/04/28/received PY - 2020/05/28/accepted PY - 2020/05/13/revised PY - 2020/7/2/pubmed PY - 2020/7/2/medline PY - 2020/7/2/entrez KW - Clinical Trial KW - Degeneration KW - Diagnostic tests/Investigation KW - Epidemiology KW - Imaging KW - Medical Education KW - Public health KW - Retina KW - Telemedicine KW - Treatment Medical JF - The British journal of ophthalmology JO - Br J Ophthalmol N2 - BACKGROUND/AIMS: Human grading of digital images from diabetic retinopathy (DR) screening programmes represents a significant challenge, due to the increasing prevalence of diabetes. We evaluate the performance of an automated artificial intelligence (AI) algorithm to triage retinal images from the English Diabetic Eye Screening Programme (DESP) into test-positive/technical failure versus test-negative, using human grading following a standard national protocol as the reference standard. METHODS: Retinal images from 30 405 consecutive screening episodes from three English DESPs were manually graded following a standard national protocol and by an automated process with machine learning enabled software, EyeArt v2.1. Screening performance (sensitivity, specificity) and diagnostic accuracy (95% CIs) were determined using human grades as the reference standard. RESULTS: Sensitivity (95% CIs) of EyeArt was 95.7% (94.8% to 96.5%) for referable retinopathy (human graded ungradable, referable maculopathy, moderate-to-severe non-proliferative or proliferative). This comprises sensitivities of 98.3% (97.3% to 98.9%) for mild-to-moderate non-proliferative retinopathy with referable maculopathy, 100% (98.7%,100%) for moderate-to-severe non-proliferative retinopathy and 100% (97.9%,100%) for proliferative disease. EyeArt agreed with the human grade of no retinopathy (specificity) in 68% (67% to 69%), with a specificity of 54.0% (53.4% to 54.5%) when combined with non-referable retinopathy. CONCLUSION: The algorithm demonstrated safe levels of sensitivity for high-risk retinopathy in a real-world screening service, with specificity that could halve the workload for human graders. AI machine learning and deep learning algorithms such as this can provide clinically equivalent, rapid detection of retinopathy, particularly in settings where a trained workforce is unavailable or where large-scale and rapid results are needed. SN - 1468-2079 UR - https://www.unboundmedicine.com/medline/citation/32606081/Prospective_evaluation_of_an_artificial_intelligence-enabled_algorithm_for_automated_diabetic_retinopathy_screening_of_30 000_patients L2 - http://bjo.bmj.com/cgi/pmidlookup?view=long&pmid=32606081 DB - PRIME DP - Unbound Medicine ER -
Try the Free App:
Prime PubMed app for iOS iPhone iPad
Prime PubMed app for Android
Prime PubMed is provided
free to individuals by:
Unbound Medicine.