Multimodal Machine Learning Using Visual Fields and Peripapillary Circular OCT Scans in Detection of Glaucomatous Optic Neuropathy.

PURPOSE: To develop and validate a multimodal artificial intelligence algorithm, FusionNet, using the pattern deviation probability plots from visual field (VF) reports and circular peripapillary OCT scans to detect glaucomatous optic neuropathy (GON). DESIGN: Cross-sectional study. SUBJECTS: Two thousand four hundred sixty-three pairs of VF and OCT images from 1083 patients. METHODS: FusionNet based on bimodal input of VF and OCT paired data was developed to detect GON. Visual field data were collected using the Humphrey Field Analyzer (HFA). OCT images were collected from 3 types of devices (DRI-OCT, Cirrus OCT, and Spectralis). Two thousand four hundred sixty-three pairs of VF and OCT images were divided into 4 datasets: 1567 for training (HFA and DRI-OCT), 441 for primary validation (HFA and DRI-OCT), 255 for the internal test (HFA and Cirrus OCT), and 200 for the external test set (HFA and Spectralis). GON was defined as retinal nerve fiber layer thinning with corresponding VF defects. MAIN OUTCOME MEASURES: Diagnostic performance of FusionNet compared with that of VFNet (with VF data as input) and OCTNet (with OCT data as input). RESULTS: FusionNet achieved an area under the receiver operating characteristic curve (AUC) of 0.950 (0.931-0.968) and outperformed VFNet (AUC, 0.868 [95% confidence interval (CI), 0.834-0.902]), OCTNet (AUC, 0.809 [95% CI, 0.768-0.850]), and 2 glaucoma specialists (glaucoma specialist 1: AUC, 0.882 [95% CI, 0.847-0.917]; glaucoma specialist 2: AUC, 0.883 [95% CI, 0.849-0.918]) in the primary validation set. In the internal and external test sets, the performances of FusionNet were also superior to VFNet and OCTNet (FusionNet vs VFNet vs OCTNet: internal test set 0.917 vs 0.854 vs 0.811; external test set 0.873 vs 0.772 vs 0.785). No significant difference was found between the 2 glaucoma specialists and FusionNet in the internal and external test sets, except for glaucoma specialist 2 (AUC, 0.858 [95% CI, 0.805-0.912]) in the internal test set. CONCLUSIONS: FusionNet, developed using paired VF and OCT data, demonstrated superior performance to both VFNet and OCTNet in detecting GON, suggesting that multimodal machine learning models are valuable in detecting GON.

SARS-CoV-2 and the Eye: Implications for the Retina Specialist from Human Coronavirus Outbreaks and Animal Models.

PURPOSE: The current SARS-CoV-2 pandemic has escalated rapidly since December 2019. Understanding the ophthalmic manifestations in patients and animal models of the novel coronavirus may have implications for disease surveillance. Recognition of the potential for viral transmission through the tear film has ramification for protection of patients, physicians, and the public. METHODS: Information from relevant published journal articles was surveyed using a computerized PubMed search and public health websites. We summarize current knowledge of ophthalmic manifestations of SARS-CoV-2 infection in patients and animal models, risk mitigation measures for patients and their providers, and implications for retina specialists. RESULTS: SARS-CoV-2 is efficiently transmitted among humans, and while the clinical course is mild in the majority of infected patients, severe complications including pneumonia, acute respiratory distress syndrome, and death can ensue, most often in elderly patients and individuals with co-morbidities. Conjunctivitis occurs in a small minority of patients with COVID-19 and SARS-CoV-2 RNA has been identified primarily in association with conjunctivitis. Uveitis has been observed in animal models of coronavirus infection and cotton-wool spots have been reported recently. CONCLUSION: SARS-CoV-2 and other coronaviruses have been rarely associated with conjunctivitis. The identification of SARS-CoV and SARS-CoV-2 RNA in the tear film of patients and its highly efficient transmission via respiratory aerosols supports eye protection, mask and gloves as part of infection prevention and control recommendations for retina providers. Disease surveillance during the COVID-19 pandemic outbreak may also include ongoing evaluation for uveitis and retinal disease given prior findings observed in animal models and a recent report of retinal manifestations.