A deep-learning system predicts glaucoma incidence and progression using retinal photographs.

BackgroundDeep learning has been widely used for glaucoma diagnosis. However, there is no clinically validated algorithm for glaucoma incidence and progression prediction. This study aims to develop a clinically feasible deep-learning system for predicting and stratifying the risk of glaucoma onset and progression based on color fundus photographs (CFPs), with clinical validation of performance in external population cohorts.MethodsWe established data sets of CFPs and visual fields collected from longitudinal cohorts. The mean follow-up duration was 3 to 5 years across the data sets. Artificial intelligence (AI) models were developed to predict future glaucoma incidence and progression based on the CFPs of 17,497 eyes in 9346 patients. The area under the receiver operating characteristic (AUROC) curve, sensitivity, and specificity of the AI models were calculated with reference to the labels provided by experienced ophthalmologists. Incidence and progression of glaucoma were determined based on longitudinal CFP images or visual fields, respectively.ResultsThe AI model to predict glaucoma incidence achieved an AUROC of 0.90 (0.81-0.99) in the validation set and demonstrated good generalizability, with AUROCs of 0.89 (0.83-0.95) and 0.88 (0.79-0.97) in external test sets 1 and 2, respectively. The AI model to predict glaucoma progression achieved an AUROC of 0.91 (0.88-0.94) in the validation set, and also demonstrated outstanding predictive performance with AUROCs of 0.87 (0.81-0.92) and 0.88 (0.83-0.94) in external test sets 1 and 2, respectively.ConclusionOur study demonstrates the feasibility of deep-learning algorithms in the early detection and prediction of glaucoma progression.FUNDINGNational Natural Science Foundation of China (NSFC); the High-level Hospital Construction Project, Zhongshan Ophthalmic Center, Sun Yat-sen University; the Science and Technology Program of Guangzhou, China (2021), the Science and Technology Development Fund (FDCT) of Macau, and FDCT-NSFC.

Assessment of Artifacts in Swept-Source Optical Coherence Tomography Angiography for Glaucomatous and Normal Eyes.

Purpose: To evaluate the frequency of and identify the factors that influence the artifacts of swept-source optical coherence tomography angiography (SS-OCTA) in glaucomatous and normal eyes. Methods: Artifacts of OCTA images of open-angle glaucoma (OAG) and normal subjects were analyzed using SS-OCTA. Univariate and multivariate logistic regression analyses were performed to evaluate the association of age, sex, best-corrected visual acuity, axial length (AL), intraocular pressure, presence and severity of OAG, and image quality score (IQS) with the presence of artifacts. Results: Images from 4426 subjects were included in the study. At least one type of artifact was present in 24.54% of the images. The most common artifacts were occurrence of motion (705 eyes, 15.93%), followed by defocus (628 eyes, 14.19%), decentration (134 eyes, 3.03%), masking (62 eyes,1.40%), and segmentation errors (23 eyes, 0.52%). Multivariate logistic analyses showed that the presence of OAG (odds ratio [OR] = 2.71; 95% confidence interval [CI], 2.09-3.51; P < 0.001), female sex (OR = 1.34; 95% CI, 1.12-1.61; P = 0.001), longer AL (OR = 1.09; 95% CI, 1.02-1.17; P = 0.017), and IQS < 40 (OR = 3.75; 95% CI, 3.15-4.48; P < 0.001) were significantly associated with higher odds for the presence of any artifact. The IQS had poor performance for detecting artifacts, with an area under the curve of 0.723, sensitivity of 73.04%, and specificity of 62.53%. Conclusions: OAG eyes had more SS-OCTA image artifacts than normal eyes. IQS is an imperfect tool for identifying artifacts. Translational Relevance: Special attention should be paid to the effect of artifacts when using SS-OCTA in the clinical setting to assess vascular parameters in patients with glaucoma.

Longitudinal Macular Retinal and Choroidal Microvasculature Changes in High Myopia.

Purpose: The purpose of this study was to determine the longitudinal changes in macular retinal and choroidal microvasculature in normal healthy and highly myopic eyes. Methods: Seventy-one eyes, including 32 eyes with high myopia and 39 healthy control eyes, followed for at least 12 months and examined using optical coherence tomography angiography imaging in at least 3 visits, were included in this study. Fovea-centered 6 × 6 mm scans were performed to measure capillary density (CD) of the superficial capillary plexus (SCP), deep capillary plexus (DCP), and choriocapillaris (CC). The rates of CD changes in both groups were estimated using a linear mixed model. Results: Over a mean 14-month follow-up period, highly myopic eyes exhibited a faster rate of whole image CD (wiCD) loss (-1.44%/year vs. -0.11%/year, P = 0.001) and CD loss in the outer ring of the DCP (-1.67%/year vs. -0.14%/year, P < 0.001) than healthy eyes. In multivariate regression analysis, baseline axial length (AL) was negatively correlated with the rate of wiCD loss (estimate = -0.27, 95% confidence interval [CI] = -0.48 to -0.06, P = 0.012) and CD loss in the outer ring (estimate = -0.33, 95% CI = -0.56 to -0.11, P = 0.005), of the DCP. The CD reduction rates in the SCP and CC were comparable in both groups (all P values > 0.05). Conclusions: The rate of CD loss in the DCP is significantly faster in highly myopic eyes than in healthy eyes and is related to baseline AL. The CD in the outer ring reduces faster in eyes with longer baseline AL.

Development and clinical deployment of a smartphone-based visual field deep learning system for glaucoma detection.

By 2040, ~100 million people will have glaucoma. To date, there are a lack of high-efficiency glaucoma diagnostic tools based on visual fields (VFs). Herein, we develop and evaluate the performance of 'iGlaucoma', a smartphone application-based deep learning system (DLS) in detecting glaucomatous VF changes. A total of 1,614,808 data points of 10,784 VFs (5542 patients) from seven centers in China were included in this study, divided over two phases. In Phase I, 1,581,060 data points from 10,135 VFs of 5105 patients were included to train (8424 VFs), validate (598 VFs) and test (3 independent test sets-200, 406, 507 samples) the diagnostic performance of the DLS. In Phase II, using the same DLS, iGlaucoma cloud-based application further tested on 33,748 data points from 649 VFs of 437 patients from three glaucoma clinics. With reference to three experienced expert glaucomatologists, the diagnostic performance (area under curve [AUC], sensitivity and specificity) of the DLS and six ophthalmologists were evaluated in detecting glaucoma. In Phase I, the DLS outperformed all six ophthalmologists in the three test sets (AUC of 0.834-0.877, with a sensitivity of 0.831-0.922 and a specificity of 0.676-0.709). In Phase II, iGlaucoma had 0.99 accuracy in recognizing different patterns in pattern deviation probability plots region, with corresponding AUC, sensitivity and specificity of 0.966 (0.953-0.979), 0.954 (0.930-0.977), and 0.873 (0.838-0.908), respectively. The 'iGlaucoma' is a clinically effective glaucoma diagnostic tool to detect glaucoma from humphrey VFs, although the target population will need to be carefully identified with glaucoma expertise input.