Diabetic Retinopathy Screening Using Smartphone-Based Fundus Photography and Deep-Learning Artificial Intelligence in the Yucatan Peninsula: A Field Study.

BACKGROUND: To compare the performance of Medios (offline) and EyeArt (online) artificial intelligence (AI) algorithms for detecting diabetic retinopathy (DR) on images captured using fundus-on-smartphone photography in a remote outreach field setting. METHODS: In June, 2019 in the Yucatan Peninsula, 248 patients, many of whom had chronic visual impairment, were screened for DR using two portable Remidio fundus-on-phone cameras, and 2130 images obtained were analyzed, retrospectively, by Medios and EyeArt. Screening performance metrics also were determined retrospectively using masked image analysis combined with clinical examination results as the reference standard. RESULTS: A total of 129 patients were determined to have some level of DR; 119 patients had no DR. Medios was capable of evaluating every patient with a sensitivity (95% confidence intervals [CIs]) of 94% (88%-97%) and specificity of 94% (88%-98%). Owing primarily to photographer error, EyeArt evaluated 156 patients with a sensitivity of 94% (86%-98%) and specificity of 86% (77%-93%). In a head-to-head comparison of 110 patients, the sensitivities of Medios and EyeArt were 99% (93%-100%) and 95% (87%-99%). The specificities for both were 88% (73%-97%). CONCLUSIONS: Medios and EyeArt AI algorithms demonstrated high levels of sensitivity and specificity for detecting DR when applied in this real-world field setting. Both programs should be considered in remote, large-scale DR screening campaigns where immediate results are desirable, and in the case of EyeArt, online access is possible.

Analysis of the Long-term Visual Outcomes of ForeseeHome Remote Telemonitoring: The ALOFT Study.

PURPOSE: To evaluate long-term visual acuity (VA) and performance of a monitoring strategy with a self-operated artificial-intelligence-enabled home monitoring system in conjunction with standard care for early detection of neovascular age-related macular degeneration (nAMD). DESIGN: Retrospective review. SUBJECTS: Patients with dry-age-related macular degeneration from 5 referral clinics. METHODS: Clinical data of patients monitored with ForeseeHome (FSH) device from August 2010 to July 2020 were reviewed. MAIN OUTCOME MEASURES: Visual acuity at baseline, VA at diagnosis of nAMD for eyes that converted while monitored, and VA from the final study follow-up, weekly frequency of use, duration of monitoring, modality of conversion diagnosis (system alert vs. detection by other standard care means), and duration and number of treatments since conversion to final study follow-up were collected. RESULTS: We reviewed 3334 eyes of 2123 patients with a mean (standard deviation [SD]) age of 74(8) years, monitored for a mean (SD) duration of 3.1 (2.4) years, with a total of 1 706 433 tests in 10 474 eye-monitoring years. The mean (SD) weekly use per patient was 5.2 (3.4), and it was persistent over the usage period. Two hundred eighty-five eyes converted while monitored at an annual rate of 2.72% and were treated with a mean (SD) 17.3 (16.5) injections over a mean (SD) 2.7 (2.0) years, with 6.4 (3.1) injections per year for eyes treated for > 1 year. The median VAs at baseline and at final follow-up for eyes that did not convert were 20/27 and 20/34 with a median change of 0.0 letters. The median VAs at baseline, conversion, and final follow-up for eyes that converted during the monitoring period were 20/30, 20/39, and 20/32 with a median change from baseline to conversion, baseline to recent, and conversion to recent of -4, -4, and 0 letters, respectively. Fifty-two percent of conversions detected had a system alert before conversion. Forty-eight percent of patients were detected by symptoms or routine visit. Patients experienced a non-nAMD alert on average every 4.6 years. At conversion and at final follow-up, the proportion (95% CI) of eyes that maintained ≥ 20/40 was 84% (78% to 88%) and 82% (76% to 86%), respectively. CONCLUSIONS: Patients in the FSH monitoring program showed excellent long-term VA years after conversion to nAMD.

Real-World Performance of a Self-Operated Home Monitoring System for Early Detection of Neovascular Age-Related Macular Degeneration.

The real-world performance of a home telemonitoring strategy (ForeseeHome AMD Monitoring System®, Notal Vision, Inc.,Manassas VA, USA) was evaluated and compared to the device arm of the AREDS2-HOME study among patients with intermediate AMD (iAMD) who converted to neovascular AMD (nAMD). All patients with confirmed conversion to nAMD who used the home monitoring system from 10/2009 through 9/2018 were identified by Notal Vision Diagnostic Clinic's medical records. Selected outcome variables were evaluated, including visual acuity (VA) at baseline and at conversion, and change in visual acuity (VA) from baseline to time of conversion. In total, 8991 patients performed 3,200,999 tests at a frequency of 5.6 ± 3.2 times/week. The 306 eyes that converted from iAMD to nAMD over the study period (a 2.7% annual rate) were included in the analyses. There was a median (interquartile range) change of -3.0 (0.0-(-10.0)) letters among converted eyes, 81% [95% confidence interval (72-88%)] maintained a VA ≥ 20/40 at the time of conversion, while 69% of the conversion detections were triggered by system alerts. The real-world performance of an at-home testing strategy was similar to that reported for the device arm of the AREDS2-HOME study. The home telemonitoring system can markedly increase early detection of conversion to nAMD.

Evaluation of a combination digital retinal camera with spectral-domain optical coherence tomography (SD-OCT) that might be used for the screening of diabetic retinopathy with telemedicine: A pilot study.

AIMS: Pilot study to determine whether an instrument combining a non-mydriatic retinal camera and spectral domain optical coherence tomography (SD-OCT) is effective for screening patients with diabetic retinopathy (DR). METHODS: Case series conducted between 2012 and 2013. DR imaged with a retinal camera/SD-OCT instrument viewed remotely was compared to a dilated examination by a retina specialist. RESULTS: The combination instrument was better than the retina specialist in detecting more severe retinopathy, primarily because of the SD-OCT. For severe retinopathy (grade ≥ 3), the image grader had better sensitivity (87.3% [95% CI: 75.5%, 94.7%]) than the retina examiner (76.4% [95% CI: 63.0%, 86.8%]). Specificities were similar between the instrument grader (96.0% [95% CI: 86.3%, 99.5%]) and retina examiner (100.0% [95% CI: 92.9%, 100.0%]). When identifying diabetic macular edema (ME), the retina examiner only identified 47.6% (20/42) of eyes with ME detected by SD-OCT. The instrument was better than a dilated retinal examination in detecting ME and not as good at detecting mild or proliferative retinopathy. CONCLUSIONS: As used in this study, the instrument was more effective in identifying DR than was the current recommendation of a dilated and comprehensive eye examination. SD-OCT is needed to accurately identify DR in a screening setting.