Purpose: The diagnosis of fungal keratitis using potassium hydroxide (KOH) smears of corneal scrapings enables initiation of the correct antimicrobial therapy at the point-of-care but requires time-consuming manual examination and expertise. This study evaluates the efficacy of a deep learning framework, dual stream multiple instance learning (DSMIL), in automating the analysis of whole slide imaging (WSI) of KOH smears for rapid and accurate detection of fungal infections. Design: Retrospective observational study. Participants: Corneal scrapings from 568 patients with suspected fungal keratitis; 51% contained filamentous fungi according to human expert interpretation. Methods: Dual stream multiple instance learning was employed to analyze WSI of KOH smears. Due to the extensive size of these images, often exceeding 100 000 pixels, conventional computer vision methods (e.g., convolutional neural networks) are not feasible. Dual stream multiple instance learning segments the WSI into patches for analysis, extracting relevant features from each patch and aggregating these to make a comprehensive slide-level diagnosis while generating heat maps to visualize areas contributing most to the prediction. Fivefold cross-validation was used for training and validation, with a hold-out test set comprising 15% of the total samples. Main Outcome Measures: Accuracy, sensitivity, specificity, area under the receiver operating characteristic curve (AUC), F1 score, positive predictive value (PPV), and negative predictive value (NPV) in distinguishing fungal from nonfungal slides. Results: Dual stream multiple instance learning demonstrated an overall AUC of 0.88 with an accuracy of 79% and an F1 score of 0.79 in distinguishing fungal from nonfungal slides, with sensitivity of 85%, specificity of 71%, PPV of 80%, and NPV of 79%. For "consensus cases," where 2 human graders agreed on the slide interpretation, the model achieved an accuracy of 85% and an F1 score of 0.85. For "discrepant cases," the accuracy was 71% with an F1 score of 0.71. The generated heatmaps highlighted regions corresponding to fungal elements. Code and models are open-sourced and available at https://github.com/Redd-Cornea-AI/KOHSmear-DSMIL. Conclusions: The DSMIL framework shows significant promise in automating interpretation of KOH smears. Its capability to handle large, high-resolution WSI data and accurately detect fungal infections, while providing visual explanations through heatmaps, could enhance the scalability of KOH smear interpretation, ultimately reducing the global burden of blindness from infectious keratitis. Financial Disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article. Ophthalmology Science 2025;5:100653 (c) 2024 by the American Academy of Ophthalmology. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).

Abstract: As artificial intelligence continues to evolve at a rapid pace, there is growing enthusiasm surrounding the potential for novel applications in corneal imaging. This article provides an overview of the potential for such applications, as well as the barriers we must overcome to realize it.

Objective: This study develops and evaluates multimodal machine learning models for differentiating bacterial and fungal keratitis using a prospective representative dataset from South India. Design: Machine learning classifier training and validation study. Participants: Five hundred ninety-nine subjects diagnosed with acute infectious keratitis at Aravind Eye Hospital in Madurai, India. Methods: We developed and compared 3 prediction models to distinguish bacterial and fungal keratitis using a prospective, consecutively-collected, representative dataset gathered over a full calendar year (the MADURAI dataset). These models included a clinical data model, a computer vision model using the EfficientNet architecture, and a multimodal model combining both imaging and clinical data. We partitioned the MADURAI dataset into 70% train/validation and 30% test sets. Model training was performed with fivefold cross-validation. We also compared the performance of the MADURAI-trained computer vision model against a model with identical architecture but trained on a preexisting dataset collated from multiple prior bacterial and fungal keratitis randomized clinical trials (RCTs) (the RCT-trained computer vision model). Main Outcome Measures: The primary evaluation metric was the area under the precision-recall curve (AUPRC). Secondary metrics included area under the receiver operating characteristic curve (AUROC), accuracy, and F1 score. Results: The MADURAI-trained computer vision model outperformed the clinical data model and the RCT-trained computer vision model on the hold-out test set, with an AUPRC 0.94 (95% confidence interval: 0.92-0.96), AUROC 0.81 (0.76-0.85), accuracy 77%, and F1 score 0.85. The multimodal model did not substantially improve performance compared with the computer vision model. Conclusions: The best-performing machine learning classifier for infectious keratitis was a computer vision model trained using the MADURAI dataset. These findings suggest that image-based deep learning could significantly enhance diagnostic capabilities for infectious keratitis and emphasize the importance of using prospective, consecutively-collected, representative data for machine learning model training and evaluation.

PURPOSE:Trachoma surveys are used to estimate the prevalence of trachomatous inflammation-follicular (TF) to guide mass antibiotic distribution. These surveys currently rely on human graders, introducing a significant resource burden and potential for human error. This study describes the development and evaluation of machine learning models intended to reduce cost and improve reliability of these surveys. METHODS:Fifty-six thousand seven hundred twenty-five everted eyelid photographs were obtained from 11,358 children of age 0 to 9 years in a single trachoma-endemic region of Ethiopia over a 3-year period. Expert graders reviewed all images from each examination to determine the estimated number of tarsal conjunctival follicles and the degree of trachomatous inflammation-intense. The median estimate of the 3 grader groups was used as the ground truth to train a MobileNetV3 large deep convolutional neural network to detect cases with TF. RESULTS:The classification model predicted a TF prevalence of 32%, which was not significantly different from the human consensus estimate (30%; 95% confidence interval of difference, -2 to +4%). The model had an area under the receiver operating characteristic curve of 0.943, F1 score of 0.923, 88% accuracy, 83% sensitivity, and 91% specificity. The area under the receiver operating characteristic curve increased to 0.995 when interpreting nonborderline cases of TF. CONCLUSIONS:Deep convolutional neural network models performed well at classifying TF and detecting the number of follicles evident in conjunctival photographs. Implementation of similar models may enable accurate, efficient, large-scale trachoma screening. Further validation in diverse populations with varying TF prevalence is needed before implementation at scale.

A real-time line-field optical coherence tomography (LF-OCT) system is demonstrated with image acquisition rates of up to 5000 B-frames or 2.5 million A-lines per second for 500 A-lines per B-frame. The system uses a high-speed low-cost camera to achieve continuous data transfer rates required for real-time imaging, allowing the evaluation of future applications in clinical or intraoperative environments. The light source is an 840 nm super-luminescent diode. Leveraging parallel computing with GPU and high speed CoaXPress data transfer interface, we were able to acquire, process, and display OCT data with low latency. The studied system uses anamorphic beam shaping in the detector arm, optimizing the field of view and sensitivity for imaging biological tissue at cellular resolution. The lateral and axial resolution measured in air were 1.7 µm and 6.3 µm, respectively. Experimental results demonstrate real-time inspection of the trabecular meshwork and Schlemm’s canal on ex vivo corneoscleral wedges and real-time imaging of endothelial cells of human subjects in vivo .

We demonstrate a new tomographic imaging device for 2D cross-sectional and 3D volumetric images of the anterior eye segment. The optical system combines optical sectioning with a stereoscopic camera setup to determine the true spatial location of the ocular structures. Moreover, we propose a calibration procedure to estimate the triangulation function that links the stereo camera pixel indices with the eye-referenced 3D spatial coordinates. The proposed instrument will image the entire anterior chamber and contiguous structures, including the corneal limbus, sclera and eyelids. Furthermore, the system can obtain corneal topography. Finally, we studied the repeatability of the instrument and its agreement with a commercial Scheimpflug imaging tomographer in a group of healthy subjects.

PURPOSE:This study sought to identify the sources of differential performance and misclassification error among local (Indian) and external (non-Indian) corneal specialists in identifying bacterial and fungal keratitis based on corneal photography. METHODS:This study is a secondary analysis of survey data assessing the ability of corneal specialists to identify acute bacterial versus fungal keratitis by using corneal photography. One-hundred images of 100 eyes from 100 patients with acute bacterial or fungal keratitis in South India were previously presented to an international cohort of cornea specialists for interpretation over the span of April to July 2021. Each expert provided a predicted probability that the ulcer was either bacterial or fungal. Using these data, we performed multivariable linear regression to identify factors predictive of expert performance, accounting for primary practice location and surrogate measures to infer local fungal ulcer prevalence, including locality, latitude, and dew point. In addition, Brier score decomposition was used to determine experts' reliability ("calibration") and resolution ("boldness") and were compared between local (Indian) and external (non-Indian) experts. RESULTS:Sixty-six experts from 16 countries participated. Indian practice location was the only independently significant predictor of performance in multivariable linear regression. Resolution among Indian experts was significantly better (0.08) than among non-Indian experts (0.01; P < 0.001), indicating greater confidence in their predictions. There was no significant difference in reliability between the two groups ( P = 0.40). CONCLUSION:Local cornea experts outperformed their international counterparts independent of regional variability in tropical risk factors for fungal keratitis. This may be explained by regional characteristics of infectious ulcers with which local corneal specialists are familiar.

PURPOSE:Landolt ring-shaped epithelial keratopathy is a corneal disease that has only been reported in 11 Japanese patients in 2014. 1 We describe the first case of Landolt ring-shaped epithelial keratopathy in the United States in a patient of European background. METHODS:This is a single case report with longitudinal care. RESULTS:A 35-year-old White patient presented with a history of ocular burning, photophobia, and decreased vision. Corneal examination showed bilateral and asymmetric microcystic lesions in a unique Landolt ring (or the letter "C") shape, distributed randomly in the epithelium. Confocal microscopy revealed cellular ballooning and hyperreflective opacities in the basal layer of the corneal epithelium. The patient has had multiple recurrences of her symptoms year-round, each lasting 4 to 8 days. Topical treatment with cyclosporine, steroids, and lubrication resolved her symptoms but without complete resolution of signs on examination. CONCLUSIONS:Our patient's clinical signs and symptoms are similar to those described previously in 11 Japanese patients. However, unlike those patients, our patient demonstrates symptomatic response to topical treatment, no seasonal association to her condition, and to date, incomplete resolution of her disease after more than 2 years. This case highlights that Landolt ring-shaped epithelial keratopathy, a novel corneal disease of unclear origin, has relevance outside of the Japanese population.

Purpose: We describe a case of metastatic conjunctival squamous cell carcinoma (SCC) presenting as an infectious sclerokeratouveitis in a patient with autoimmune disease. Observations: A 63-year-old Caucasian female presented to the cornea service with a raised perilimbal scleral infiltrate, hypopyon, and corneal perforation concerning for infectious necrotizing sclerokeratoveitis. She had an ocular history of a recurrent “pyogenic granuloma” and her medical history was notable for well-controlled systemic lupus erythematosus and rheumatoid arthritis. Scleral debridement and repair with a corneal patch graft was performed. Bacterial cultures grew 4+ Proteus mirabilis sensitive to tobramycin. One month later, she developed bulky painless scleral lesions with leukoplakic features. A scleral biopsy revealed squamous cell carcinoma. Imaging suggested ciliary body and medial rectus infiltration without perineural invasion. Right anterior exenteration was performed as immunotherapy was felt to be unlikely to achieve success in light of her autoimmune conditions. Six months later, she developed a palpable right parotid mass with biopsy confirming metastatic squamous cell carcinoma. She underwent a right parotidectomy and is now undergoing consolidative radiotherapy. Conclusions and Importance: Ocular surface neoplasia can present as a necrotizing sclerokeratouveitis, contributed by both the tumor and an atypical infectious process. Malignancy with superinfection should be in the differential diagnosis of recurrent ocular surface inflammation.

Importance:Infectious conjunctivitis can lead to corneal involvement and result in ocular morbidity. The identification of biomarkers associated with corneal involvement has the potential to improve patient care. Objective:To identify biomarkers in patients with acute infectious conjunctivitis. Design, Setting, and Participants:This cross-sectional study took place from December 2016 to March 2024. Analyses were performed in 3 phases. First, logistic regression and machine learning algorithms were used to predict the probability of demonstrating corneal involvement in patients with presumed infectious conjunctivitis. Second, quantitative reverse transcription polymerase chain reaction (RT-qPCR) was used to confirm the most important biomarker gene identified by the algorithm. Third, the biomarker gene was validated in prospectively collected conjunctival samples of adult patients from 3 outpatient centers in Thailand and 1 in India. Patients with signs and symptoms of infectious conjunctivitis and onset within less than 14 days were eligible. Exclusion criteria were the inability to consent, presumed toxicity, or allergic conjunctivitis. Exposures:Acute infectious conjunctivitis. Main Outcomes and Measures:The identification and validation of ocular surface gene expression associated with corneal findings on slitlamp examination. Results:Thirteen genes exhibited a 1.5-log2 fold change in expression in patients with corneal involvement compared to patients without corneal involvement. Using the 13 genes to train and cross validate, logistic regression produced the highest mean area under the receiver operating characteristic curve (AUROC; 0.85; 95% CI, 0.84-0.86) for corneal involvement. The removal of apolipoprotein E (APOE) from the gene ensemble led to a decline in predictive performance of the logistic regression classifier (from mean AUROC 0.85 [95% CI, 0.84-0.86] to 0.74 [95% CI, 0.73-0.75]; adjusted P = .001 [Tukey test]). Orthogonal testing of APOE expression level with RT-qPCR showed that APOE expression was higher in patients with corneal involvement compared to patients without (median [IQR], 0.23 [0.04-0.47] vs 0.04 [0.02-0.06]; P = .004 [Mann-Whitney U test]). Using a Youden index of 0.23 Δ threshold cycle, APOE had a sensitivity of 56% (95% CI, 33-77) and a specificity of 88% (95% CI, 79-93) in 106 samples with conjunctivitis at Aravind, India (P < .001 [Fisher exact test]). When applied to a different patient population in Thailand, the same criteria could discriminate between disease states (58 samples; sensitivity, 47%; 95% CI, 30-64 and specificity, 93%; 95% CI, 77-99; P = .001 [Fisher exact test]). Conclusions and Relevance:The results from this study suggest that the host conjunctival immune response can be meaningfully interrogated to identify biomarkers for ocular surface diseases.

Objective: To investigate the impact of corneal photograph quality on convolutional neural network (CNN) predictions. Design: A CNN trained to classify bacterial and fungal keratitis was evaluated using photographs of ulcers labeled according to 5 corneal image quality parameters: eccentric gaze direction, abnormal eyelid position, over/under-exposure, inadequate focus, and malpositioned light reflection. Participants: All eligible subjects with culture and stain-proven bacterial and/or fungal ulcers presenting to Aravind Eye Hospital in Madurai, India, between January 1, 2021 and December 31, 2021. Methods: Convolutional neural network classification performance was compared for each quality parameter, and gradient class activation heatmaps were generated to visualize regions of highest influence on CNN predictions. Main Outcome Measures: Area under the receiver operating characteristic and precision recall curves were calculated to quantify model performance. Bootstrapped confidence intervals were used for statistical comparisons. Logistic loss was calculated to measure individual prediction accuracy. Results: Individual presence of either light reflection or eyelids obscuring the corneal surface was associated with significantly higher CNN performance. No other quality parameter significantly influenced CNN performance. Qualitative review of gradient class activation heatmaps generally revealed the infiltrate as having the highest diagnostic relevance. Conclusions: The CNN demonstrated expert-level performance regardless of image quality. Future studies may investigate use of smartphone cameras and image sets with greater variance in image quality to further explore the influence of these parameters on model performance. Financial Disclosure(s): Proprietary or commercial disclosure may be found after the references.

Casey Eye Institute, Oregon Health & Science University, Portland, OR Financial disclosures/conflicts of interest: None reported.

Corneal disease is the fifth leading cause of global blindless. Optical coherence tomography (OCT) for anterior imaging is extensively used due to its non-invasive and high-resolution volumetric imaging characteristics. Optical coherence microscopy (OCM) is a technical variation of OCT that can image the cornea with cellular resolution. Here, we demonstrate a visible-light OCM as a low-cost and easily reproducible system to visualize various corneal cellular structures such as epithelial cells, endothelial cells, keratocytes, and collagen bundles within stromal lamellae. The visible-light OCM was also used to study pressure changes in anterior segment human donor eyes. The system achieved an axial resolution of 12 μm in tissue over a 1.2 mm imaging depth, and a lateral resolution of 1.6 µm over a field of view of up to 750 μm × 750 μm.

Purpose: The aim of this study was to describe a case of corneal involvement as an early manifestation of ocular disease in the 2022 human mpox (monkeypox) virus outbreak. Methods: This is a single case report with longitudinal care. Results: A 47-year-old immunocompetent man presented with viral conjunctivitis before development of skin lesions or systemic symptoms. Subsequently, he developed membranous keratoconjunctivitis and a corneal epithelial defect. Orthopoxvirus-positive polymerase chain reaction test from his ocular surface was positive. The epithelial defect did not heal with conservative treatment but was successfully treated with amniotic membrane transplantation over 8 days. Reduced corneal sensation was noted after epithelial healing, and polymerase chain reaction from the ocular surface remained positive at 17 days from symptom onset, with slowly recovering conjunctivitis at 21 days. Continued membrane formation required repeated removal but significantly improved with topical corticosteroid treatment after epithelial healing by 29 days of symptom onset. Corneal sensation normalized by 87 days from symptom onset at which time symblepharon were noted but PCR testing from the ocular surface was negative. Conclusions: Early corneal involvement of human monkeypox virus is possible. Transient corneal hypoesthesia may be due to acute inflammation. Chronic inflammatory changes can result in symblepharon. These findings have potential implications in patient care and corneal donation.

Supplemental Digital Content is Available in the Text. Purpose: Management of ocular rosacea is challenged by the limited evidence-based systemic treatment guidelines and lack of elucidated mechanisms of treatment efficacy. Methods: We conducted an online survey of clinicians who regularly treat ocular rosacea to elicit their opinions on treatment algorithms and understanding of the treatment's primary mechanism of action. Descriptive statistics and univariate comparisons were reported. Results: One hundred thirty-two participants completed the online survey. Of the 132 respondents, 74% were cornea specialists. Most respondents (85%) favored systemic tetracyclines over macrolides. Providers' specialty training did not significantly influence preference between tetracyclines and macrolides for ocular rosacea management. Among tetracycline prescribers, there was no consensus regarding initial dosage and duration prescribing patterns. Most macrolide prescribers (88%) initiated a 3-week course of 1 gram of azithromycin weekly. Long-term management strategy for treatment-responsive patients varied: 46% preferred to half the initial dose, 29% discontinued pharmacotherapy, and 16% chronically pulse-dosed patients. Most tetracycline prescribers (90%) and macrolide prescribers (73%) postulate their chosen agents' primary mechanism of effect for ocular rosacea is anti-inflammatory. However, there was no consensus in identifying anti-inflammatory doses of either drug class. Furthermore, there is discordance between prescribers’ intended mechanistic effect with the selection of initial dosages for both tetracycline and macrolides for ocular rosacea. Conclusions: Among clinicians who commonly treat ocular rosacea, there is significant community equipoise regarding which dose of tetracycline is best for initial systemic treatment of this disease. In addition, a consensus understanding regarding mechanism of action of this treatment is lacking.

Purpose:To report and present images of a case in which discrete conjunctival lesions developed in the setting of primary varicella zoster virus infection (ie, chickenpox).Methods:Case report and literature review.Results:This report describes a young, unvaccinated male who developed an acutely painful, red eye in the setting of disseminated primary varicella zoster infection. The cutaneous rash was widespread and included lesions on both eyelids. The patient was found to have multiple discrete de-epithelialized lesions involving the palpebral and bulbar conjunctiva. Throughout the disease course, good visual function was maintained and there was no evidence of intraocular involvement. The ocular surface lesions resolved without sequelae after 1 week of treatment with topical antibiotic ointment.Conclusions:Primary varicella zoster infection is an increasingly rare phenomenon in the setting of widespread vaccination. However, unvaccinated or undervaccinated individuals and other at-risk populations remain susceptible to developing severe infections. This case of chickenpox involved discrete conjunctival lesions that resolved without sequelae after conservative treatment with topical antibiotic ointment. While serious ophthalmic complications are uncommon in primary varicella infection, clinicians should be aware of the potential for ocular morbidity in this increasingly rare condition.