Glaucoma is a chronic neurodegenerative disease of the visual system, and treatment is targeted toward lowering intraocular pressure. However, some patients fail to respond to treatment and their intraocular pressure levels remain high, risking continuous vision loss. Explainable machine learning provides a mechanism for both individual prognostication and the identification of factors associated with treatment outcome. Here, we used explainable machine learning to predict intraocular pressure for glaucoma patients receiving medication treatment. We accessed the UK Biobank to obtain information on 290 eyes from 161 participants who reported a diagnosis of glaucoma and were receiving treatment. Features were divided into three distinct datasets containing demographic data only, physiometabolic parameters and medication prescription data, and all data combined. We evaluated five machine learning techniques for each feature set in terms of their ability to predict intraocular pressure at a follow-up visit in a classification task. We then calculated SHapley Additive exPlanation (SHAP) values for the best performing model to determine feature importance, stability, and interactions. We found that eXtreme Gradient Boosting (XGBoost) outperformed all other models when trained and tested on the combined feature set with an area under receiver operating characteristic curve (AUC) of 0.708. Insulin-like growth factor 1 (IGF-1), low-density lipoprotein (LDL), and lymphocyte count ranked as the three most important features for this model. LDL and IGF-1 exhibited a low degree of global variability in contribution to the model output across all cross-validation repeats. SHAP values demonstrated the strongest interactions being between LDL and IGF-1. In summary, our studies indicated the importance of blood LDL and IGF-1 in contributing to the outcomes of intraocular pressure lowering treatment and demonstrated the ability of XGBoost to predict these outcomes.

Visual impairment affects over 250 million people globally and carries a significant psychosocial and economic burden. In recent years, rapid advancements in biomedical engineering and regenerative medicine have driven the development of innovative therapies aimed at restoring visual function. Here, we first provide an overview of the visual pathway, imaging modalities, and the psychosocial and economic burden of vision loss. Then we review four vision restoration approaches: gene and stem cell therapies, optogenetics, retinal prosthetic devices, and visual pathway electrical stimulation. For each therapeutic approach, the mechanism of action, clinical outcomes, adverse events, and current limitations are discussed. Stem cell transplantation and gene-editing technologies are being explored to repair or replace damaged retinal cells at the molecular level. Retinal prosthetics implanted in various layers of the retina induce visual experiences in patients with degenerative diseases, though challenges in resolution and long-term performance remain. Lastly, electrical stimulation techniques, including transcorneal and transorbital stimulation, have shown potential to enhance residual vision through neuroprotective measures and neuroplastic synchronization of neuronal signaling activity. This article reviews and compares these therapies to give a balanced perspective on treatment modalities for visual impairment in a range of retinal diseases and optic neuropathies.

Cerebrospinal fluid (CSF), partly driven by sensory stimulation, is crucial for maintaining homeostasis and clearing metabolic waste. Whether such stimulus-driven CSF flow is disrupted in age-related neurodegenerative diseases of the visual system remains unclear. This study examined the CSF flow during visual stimulation in glaucoma patients and healthy older adults using functional magnetic resonance imaging. In glaucoma, CSF inflow becomes progressively decoupled from the visually evoked blood-oxygenation-level-dependent (BOLD) response. Specifically, the characteristic stimulus-locked CSF patterns, which decrease after stimulus onset and increase after offset, diminish with disease severity. Mediation analysis suggests this flattened CSF pattern is driven by a flatter ascending BOLD slope, leading to a shallower CSF trough and a reduced post-stimulus surge. These results indicate that glaucoma-related functional impairments contribute to downstream alterations in CSF dynamics. Overall, this study provides insight into how glaucoma disrupts visually driven CSF inflow and highlights in vivo biomarkers for monitoring CSF dynamics.

Objective: To evaluate the impact of training and testing deep learning (DL) models for visual field (VF) forecasting using input-target pairs in which the target is either the measured VF test result, or its smoothed counterpart constructed via linear regression. Design: A retrospective data analysis study evaluating DL models for VF forecasting under multiple training and testing configurations. Subjects and Controls: A total of 1400 subjects (healthy and glaucoma patients) with 19 437 reliable Humphrey VF (24-2 Swedish Interactive Threshold Algorithm) tests collected from longitudinal cohorts at the University of Pittsburgh and New York University. Methods: Three DL-based pointwise VF forecasting methods were trained and tested under 4 different configurations formed by using measured and smoothed VF targets. Smoothed targets were constructed by applying linear regression over triplets of consecutive VF tests. Models were assessed using fivefold cross-validation and mean absolute error (MAE) as the training and testing metric. Main Outcome Measures: Mean absolute error of forecasted VF test results under various training and testing configurations. Results: Models trained and tested on smoothed VF targets consistently achieved lower MAEs compared to those trained and tested on measured VF targets. The performance improvements were most prominent in the 0.5- to 1.5-year forecast range. Furthermore, models trained with smoothed VF targets showed comparable performance when evaluated against measured VF targets. Conclusions: Using smoothed VF targets for training improves forecasting accuracy by guiding DL models to learn long-term trends in the data rather than forcing them to model noise and short-term variabilities, which are prevalent in VF test data. This approach aligns with clinical goals of assessing meaningful functional changes over time and suggested to be considered in future DL-based VF modeling efforts. Financial Disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

There has been a tremendous amount of image processing and machine learning research to measure and classify disease progression from live optical coherence tomography (OCT) imaging of the retina. The images considered here are large, complex, three-dimensional (3-D) and difficult to visualize effectively. Many current supervised machine learning approaches, e.g. neural networks, are non-metric meaning that any features or measurements generated can introduce systematic distortion that may be correlated with underlying non-meaningful physiological differences. Here we present a metric learning approach using the normalized compression distance (NCD) combined with anisotropic structure-enhancing filters to quantify and visualize the principal differences among a collection of 3-D retinal images. We validate the NCD-measured structural differences between pairs of images against the physician-measured change in visual field function, achieving a prediction error of ∼ 0.5 dB, more accurate than non-metric deep learning approaches. The normalized compression vectors (NCV) are proposed as a feature set measuring visual differences among a collection of 3-D microscopy images. The utility of the NCV for visualizing and measuring patterns of change is demonstrated for a human with moderate non-progressing glaucoma and for a non-human primate model using intraocular pressure setting manipulation. We conclude with a brief simulation of non-metric embedding features, e.g. from neural networks, introducing class-correlated statistical distortion.

The diagnosis and monitoring of glaucoma require precise evaluation of ocular structural features. The advent of ocular imaging has revolutionized both the clinical management and research of glaucoma, establishing itself as a cornerstone of contemporary practice. In this review, we summarize the major advances in ocular imaging technologies and their contributions to the understanding, diagnosis, and monitoring of glaucoma over the past 2 centuries. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Non-human primates (NHPs) are a crucial model for studying glaucoma because of their similarities to humans in anatomy, physiology and pathology. Our goal in this study was to quantify in vivo NHP lamina cribrosa (LC) shapes at low, normal, and elevated intraocular pressures (IOPs), and compare them with literature reported values for in vivo human LCs. We imaged the optic nerve heads (ONH) of seven eyes from six rhesus macaque monkeys using spectral-domain optical coherence tomography (SD-OCT) at several IOP levels while keeping intracranial pressure at baseline. LC shape was characterized by measuring shape index (SI) and curvedness from manually delineated marks of the anterior LC surface. We found that LC shape in NHPs was more similar across individuals at normal IOP than it is in humans. This was, in part, because the NHPs LCs did not exhibit the characteristic horizontal central ridge seen in human LCs and were instead less saddle-shaped and more troughshaped. With increasing IOP, NHPs LCs curvedness increased, without significant changes in SI, differing from the human response where SI decreases. These findings emphasize the importance of characterizing speciesspecific differences in anatomy and biomechanics, and the need to determine how these differences may impact susceptibility to glaucomatous damage.

Purpose: To determine the efficacy and safety of repetitive transorbital alternating current stimulation (rtACS) treatment by assessing vision-related quality of life and visual function outcome in subjects treated with rtACS versus sham-control. Study design: Double masked, randomized, sham-controlled clinical trial (NCT03188042). Subjects: Sixteen subjects with moderate-to-advanced glaucoma (visual field [VF] mean deviation [MD] <=-6.00 decibels) randomized into sham (9 subjects) or rtACS intervention (7 subjects) groups. Methods: Subjects underwent 10 rtACS sessions over 2 weeks. All subjects had comprehensive ocular examination at baseline, 1-week, and 4-weeks posttreatment. Main Outcome Measures: Visual acuity (VA), contrast sensitivity (CS), VF MD, number of threshold sensitivity points that changed or were unchanged, and vision-related quality of life (VR-QoL) questionnaire scores. Results: The rtACS group showed a significantly greater improvement from baseline to 4 weeks posttreatment compared with sham in VR-QoL domains including near activities (P < 0.01), dependency (P = 0.03), social functioning (P = 0.03), mental health (P < 0.01) and in the overall composite score (P = 0.04). No significant changes were detected with VA, CS, and VF analyses for either group. No serious adverse events were noted in either study group. Conclusions: Repetitive transorbital alternating current stimulation therapy showed a significant beneficial effect on several domains of VR-QoL. Further studies will determine its utility in glaucoma.

Currently there are no surgical solutions to restore vision in the irreversibly blind. Whole eye transplantation (WET), is an appealing surgical approach for restoration, replacement, and reconstruction of nonfunctioning eyes. Development of a reliable animal model to test the integrity and functionality of the transplanted eye is an essential step towards clinical whole eye transplantation. This study presents a feasible vascularized orthotopic eye transplantation preclinical rat model to study the structural and functional outcomes of whole eye transplantation. Syngeneic orthotopic transplants were performed in rats, involving anastomoses between carotid arteries, external jugular veins, and optic nerve coaptations of donors and recipients. The transplanted and recipient native eyes were assessed by ocular exam under anesthesia, optical coherence tomography (OCT), histology, magnetic resonance imaging and electroretinography. A 100% surgical survival rate of recipients with maintained long-term health demonstrated this to be a reliable and reproducible model. Assessment from clinical examination under anesthesia revealed that segments of native eyes appeared normal throughout the duration of the study, but transplanted eyes presented mild chemosis of the eye lids, mild ciliary flush of the conjunctiva, cornea neovascularization, mild engorgement of the vessels in the iris, and mild opacities in the lens in some animals. Most of these findings improved over time after transplantation. Doppler optical coherence tomography corroborated the presence of blood flow in transplanted retinas. There was no significant difference in measured IOP between native and transplanted eyes. Both histology and OCT scans demonstrated increased central corneal thickness and decreased total retinal thickness in transplanted eyes. Transplanted eyes exhibit minimal scotopic and photopic ERG responses. To date, no other vascularized orthotopic rodent WET transplantation models have been described in the literature. As functional visual return remains the ultimate goal, this model provides a foundation for future translational strategies and is ideal for testing immunomodulatory, neuroprotective, and neuroregenerative approaches either individually or in combination, as required for total human eye allotransplantation (THEA) to become a clinical reality.

Purpose:The integration of artificial intelligence (AI), particularly deep learning (DL), with optical coherence tomography (OCT) offers significant opportunities in the diagnosis and management of glaucoma. This article explores the application of various DL models in enhancing OCT capabilities and addresses the challenges associated with their clinical implementation. Methods:A review of articles utilizing DL models was conducted, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), generative adversarial networks (GANs), autoencoders, and large language models (LLMs). Key developments and practical applications of these models in OCT image analysis were emphasized, particularly in the context of enhancing image quality, glaucoma diagnosis, and monitoring progression. Results:CNNs excel in segmenting retinal layers and detecting glaucomatous damage, whereas RNNs are effective in analyzing sequential OCT scans for disease progression. GANs enhance image quality and data augmentation, and autoencoders facilitate advanced feature extraction. LLMs show promise in integrating textual and visual data for comprehensive diagnostic assessments. Despite these advancements, challenges such as data availability, variability, potential biases, and the need for extensive validation persist. Conclusions:DL models are reshaping glaucoma management by enhancing OCT's diagnostic capabilities. However, the successful translation into clinical practice requires addressing major challenges related to data variability, biases, fairness, and model validation to ensure accurate and reliable patient care. Translational Relevance:This review bridges the gap between basic research and clinical care by demonstrating how AI, particularly DL models, can markedly enhance OCT's clinical utility in diagnosis, monitoring, and prediction, moving toward more individualized, personalized, and precise treatment strategies.

PURPOSE. Proteomes of lens nuclei from young (4 years old) and old (15-16 years old) rhesus macaques (Macaca mulatta) were analyzed to determine similarity of the proteomic profile to that of human lenses, age-related differences in protein solubility, and association of various post-translational modifications with age and protein solubility. METHODS. Lens core proteins were separated into water-soluble and water-insoluble fractions using aqueous buffer and centrifugation. The water-insoluble fraction was solubilized using sodium dodecyl sulfate (SDS). Proteins were processed using S-trap columns, and peptide digests were analyzed using high-resolution, label-free datadependent acquisition (DDA) proteomics. Open modification searches were performed using MSFragger to identify possible post-translational modifications (PTMs). The number of modified peptide tandem mass spectra confidently assigned to samples by age or solubility were compared to find PTMs with statistically significant count differences. RESULTS. The overall proteomic profile of rhesus macaque lenses was very similar to human lenses, consisting of 80.2% crystallins, 1.1% beaded filament proteins, and 18.7% other proteins. The crystallin fraction consisted of 27% alpha crystallins, 67.6% beta/gamma crystallins, and 5.4% taxon-specific psi crystallin. Glycolytic enzymes, beta/gamma crystallins, and a few glutathione-related enzymes were found to have agerelated shifts to the water-insoluble fraction. There were significant differences in deamidation, dioxidation, carbamylation, carboxymethylation, and trioxidation based on age and/or solubility of proteins. CONCLUSIONS. These data indicate a high level of conformity between rhesus macaque and human lens proteomes, and a few key differences. We identified several age-related differences in protein solubility and PTM that may contribute to lens pathology.

Purpose:Retinal nerve fiber layer thickness (RNFLT), a glaucoma biomarker, has a wide normative range affecting its sensitivity and specificity for abnormality detection. The interindividual RNFLT peak location variability contribution to this wide normative range has not been directly evaluated. The purpose of this study is to assess the effect of RNFLT peak normalization (PN) on normative variability. Methods:Circumpapillary RNFLT profiles at 1.7 mm radius from the optic nerve head (ONH) were re-sampled from optical coherence tomography (OCT) volumes (Cirrus HD-OCT, 200 × 200) obtained from one eye of 83 healthy individuals. Fovea-ONH axis (FOA) was calculated from corresponding scanning laser ophthalmoscope images. Supratemporal (ST) and infratemporal (IT) RNFLT peaks of each profile were aligned to respective average peak locations. Normative ranges were calculated by averaging individual profiles before and after PN (with and without FOA to horizontal image axis (HA) alignment). Results:RNFLT-PN resulted in an overall decrease in coefficient of variation (CoV) of the normative range by 4.2% (P = 0.02). CoV was reduced by more than 10% in clock-hours 10 (11.9%), 8 (10.6%), 6 (10.4%) after PN, and 7 (16.3%), 10 (11.4%), and 12 (10.4%) after PN with FOA-HA alignment. RNFLT-PN corrected for abnormality categorization because of peak misalignment in RNFLT profiles of healthy and glaucoma suspect subjects. Conclusions:RNFLT-PN reduces normative variability, especially in the ST and IT regions. Translational Relevance:RNFLT-PN reduces normative variability and improves sectoral abnormality categorization, potentially leading to better sensitivity and specificity of RNFLT measure in glaucoma detection.

The ocular surface microbiota (OSM) is important for eye health, and variations in OSM composition have been associated with multiple diseases in humans. Studies of OSM-disease dynamics in humans are confounded by lifestyle factors. Animal models provide a complementary approach to understanding biological systems, free from many confounds of human studies. Here, we provide the first study of the OSM of rhesus macaques, a premier animal model for eye health and disease. We describe the taxonomy of the rhesus macaque OSM, and explore compositional correlations with age, sex, and living condition. We analyzed eyelid and conjunctival microbiota swabs from 132 individual rhesus macaques (Macaca mulatta) (57 males, 75 females, 1–26 years old) from one captive and one free-ranging group using 16 S rRNA V3/V4 MiSeq sequencing. We investigated alpha diversity, beta diversity, and differential abundance. We found several similarities between the top Phyla and Genera of the rhesus macaque OSM and those reported in human literature. Significantly higher alpha diversity, which may reflect age-related ocular surface mucous membrane integrity and immune function, was present in younger individuals compared to older ones. Higher alpha diversity was also present in free-ranging rhesus macaques compared to ones in captivity, possibly related to differences in diet, exercise, and medical exposures between macaques in different living conditions. Beta diversity was most strongly influenced by individual identity, followed by living conditions. Sex did not correlate with any OSM variation. In this study we describe the taxonomic composition of the rhesus macaque OSM, and identify significant differences in alpha and beta diversity according to individual nonhuman primate host variables and the surrounding environment. Our findings suggest composition of the nonhuman primate OSM is shaped by age-related physiology, individual identity, and external living conditions. Our results offer novel insights into an underexplored region of the primate microbiome and highlight the utility of rhesus macaques as a model system for investigating the links between the OSM, ocular health, and disease.