Referral for disease-related visual impairment using retinal photograph-based deep learning: a proof-of-concept, model development study

Abstract

Background In current approaches to vision screening in the community, a simple and efficient process is needed to identify individuals who should be referred to tertiary eye care centres for vision loss related to eye diseases. The emergence of deep learning technology offers new opportunities to revolutionise this clinical referral pathway. We aimed to assess the performance of a newly developed deep learning algorithm for detection of disease-related visual impairment. MethodsIn this proof-of-concept study, using retinal fundus images from 15 175 eyes with complete data related to best-corrected visual acuity or pinhole visual acuity from the Singapore Epidemiology of Eye Diseases Study, we first developed a single-modality deep learning algorithm based on retinal photographs alone for detection of any diseaserelated visual impairment (defined as eyes from patients with major eye diseases and best-corrected visual acuity of <20/40), and moderate or worse disease-related visual impairment (eyes with disease and best-corrected visual acuity of <20/60). After development of the algorithm, we tested it internally, using a new set of 3803 eyes from the Singapore Epidemiology of Eye Diseases Study. We then tested it externally using three population-based studies (the Beijing Eye study [6239 eyes], Central India Eye and Medical study [6526 eyes], and Blue Mountains Eye Study [2002 eyes]), and two clinical studies (the Chinese University of Hong Kong's Sight Threatening Diabetic Retinopathy study [971 eyes] and the Outram Polyclinic Study [1225 eyes]). The algorithm's performance in each dataset was assessed on the basis of the area under the receiver operating characteristic curve (AUC). Findings In the internal test dataset, the AUC for detection of any disease-related visual impairment was 94•2% (95% CI 93•

Deep learning algorithms for automatic detection of pterygium using anterior segment photographs from slit-lamp and hand-held cameras

“…These considerations include, the deployment site (ie, whether in rural community screening sites or primary care facilities), local regulatory requirements for implementation of health technology (which may require minimal levels of specificity and sensitivity to be achieved before rollout is granted), and availability of healthcare resources/ facilities for treatment (ie, communities with finite resources may opt for a more stringent threshold which would yield higher positive predictive value, to more strictly identify those which truly require treatment. 34 Fourth, it should be noted that the ground truth of referable pterygium was defined based on the presence of pterygium with >2.50 mm extension towards the cornea or with a base width of >5.00 mm, but did not take total area and fleshiness (ie, thickness) 20 of the pterygium into account. However, previous studies have indicated that horizontal extension along with the base width of the pterygium [23][24][25][26] have the greatest influence on corneal astigmatism and ocular aberrations.…”

AI-automated referral for patients with visual impairment

“…The ResNet model did well in predicting the presence of any diseaserelated visual impairment, with an area under receiver operating characteristic curve (AUC) range varying from 86•6% (95% CI 83•4-89•7) to 93•6% (92•4-94•8) across the external validation datasets. 9 In external datasets, the AUC range for predicting mild diseaserelated visual impairment was 81•9% (77•2-86•6) to 92•5% (90•8-94•2) and for moderate diseaserelated visual impairment was 85•9% (81•8-90•1) to 93•5% (91•7-95•3). According to these AUC data, the model did quite well to identify disease-related visual impairment in the external datasets, indicating potential generalisability, which is a crucial element for widespread deployment and clinical utility.…”

“…In April, 2018, the US Food and Drug Administration approved the first AI-based referral device for diabetic retinopathy. 8 In The Lancet Digital Health, Yih-Chung Tham and colleagues 9 report development of a single modality (fundus photograph) deep learning algorithm for the detection of disease-related visual impairment with the goal of creating a referral system in vision care. Following the recommendation of WHO, some high-income countries have implemented annual vision screening, especially for older populations.…”

“…Following the recommendation of WHO, some high-income countries have implemented annual vision screening, especially for older populations. 9 Some screening systems rely on visual acuity checks whereas others use retinal photography to identify disease for referral to tertiary care. To incorporate visual acuity and retinal images for tertiary referral, large-scale resources and trained personnel are required.…”

Artificial intelligence applications and cataract management: A systematic review

“…79In 2021, Yih-Chung Tham and coworkers proposed a new method for disease-related visual impairment diagnosis, employing retinal photograph-based deep learning 65. Patients affected by different major age-related eye diseases such as cataract, diabetic retinopathy, and maculopathy were included, and a deep CNN was used (Residual Neural Network ResNet-50 architecture).…”