Aravind Eye Hospitals is a hospital chain in India. It was founded by Dr. Govindappa Venkataswamy (popularly known as Dr.V) at Madurai, Tamil Nadu in 1976. It has grown into a network of eye hospitals and has had a major impact in eradicating cataract related blindness in India. As of 2012, Aravind has treated nearly 32 million patients and performed 4 million surgeries. The model of Aravind Eye Care hospitals has been applauded and has become a subject for numerous case studies across the world.
This review examines the promising potential of Artificial Intelligence (AI) in pediatric ophthalmology, specifically in the assessment, prediction, management, and treatment of myopia in children. The use of AI, particularly machine learning (ML) and deep learning (DL), in predicting myopia in children has garnered significant interest for its potential in early screening, detection, prognosis prediction, monitoring of anti-myopia treatment, personalized interventions, and proactive management strategies. This review aims to summarize the current literature on myopia and AI, presenting them as emerging trends and future directions in the pediatric population, and highlighting emerging strategies for the future of myopia management.
The lack of physiologically relevant and controllable experimental systems has limited mechanistic understanding of age-related macular degeneration (AMD) and the development of effective therapeutic strategies. Here, we present a tunablein vitroretinal pigment epithelium (RPE) stress model that integrates engineered light delivery with lipid modulation to reproduce early AMD-like cellular pathology under standard culture conditions. Human RPE cells (ARPE-19 and iPSC-derived RPE) were exposed to precisely controlled, low-intensity light-induced oxidative stress in the presence of docosahexaenoic acid (DHA), a highly unsaturated retinal lipid, or palmitic acid (PA) as a saturated lipid control. Cellular responses were assessed using functional and structural readouts including lysosomal and mitochondrial activity, membrane integrity, epithelial morphology, tight junction organization, and lipid peroxidation. A programmable LED-based exposure system enabled fine control over light intensity, duration, and cycling, allowing delivery of sub-lethal, chronic oxidative stress. Combined light and DHA exposure selectively induced lipid peroxidation, disruption of ZO-1-defined tight junctions, and progressive loss of RPE viability, while PA-treated cells and non-retinal HuH7 hepatocytes showed minimal sensitivity. ARPE-19 cells responded rapidly, whereas iPSC-derived RPE cells exhibited delayed but comparable pathological changes, reflecting differences in cellular maturity and stress resilience. Pharmacological inhibition of ferroptosis using ferrostatin-1 significantly reduced lipid peroxidation and rescued epithelial integrity and cell viability, identifying ferroptosis as a key mechanism underlying RPE vulnerability in this system. By enabling programmable and reproducible delivery of oxidative lipid stress, this modular light-based platform provides a biofabrication-compatible framework for modeling early AMD, with potential for integration into more complex retinal constructs, co-culture systems, and high-throughput therapeutic screening pipelines.
Purpose:To compare the clincial setting effectiveness of three myopia control spectacle lenses, defocus incorporated multiple segments (DIMS), highly aspherical lenslet target (HALT), and cylindrical annular refractive elements (CARE), versus single-vision lenses (SVLs) in attenuating axial elongation and myopic refractive shift in children. Design:A single-center, retrospective cohort study. Participants:A total of 899 subjects (1780 eyes) aged 6-16 years with cycloplegic spherical equivalent (SE) between -0.50 D and -9.00 D. Subjects were divided into four groups based on prescribed eyewear: SVL (n = 446 eyes), DIMS (n = 523 eyes), HALT (n = 295 eyes), and CARE (n = 516 eyes). Methods:Data were extracted from electronic medical records. Longitudinal changes in axial length (AL) and SE refraction were analyzed using linear mixed-effects models to compare progression rates among the lens types, adjusting for baseline age, sex, parental myopia, and the baseline value of the respective outcome variable. Main Outcome Measures:Annualized rate of progression in AL (mm/year) and SE (D/year). Results:The mean age of our cohort was 12.18 ± 2.73 years, and the mean follow-up time was 12.49 ± 5.71 months. All myopia control lenses significantly slowed AL elongation compared to the SVL group (0.193 mm/year). The adjusted annual AL progression was 0.098 mm/year for DIMS, 0.088 mm/year for CARE, and 0.054 mm/year for HALT (P < 0.001 for all vs. SVL). Highly aspherical lenslet target lenses demonstrated significantly greater efficacy in slowing AL elongation compared to both DIMS (P < 0.001) and CARE (P value < 0.001). All myopia control lenses also significantly reduced myopic SE progression (-0.370 D/year for SVL) with rates of -0.107 D/year for DIMS, -0.106 D/year for CARE, and -0.073 D/year for HALT. However, pairwise comparisons among the three myopia control lens types for SE progression did not reach statistical significance (P > 0.13). Conclusions:Highly aspherical lenslet target, DIMS, and CARE effectively slowed progression versus SVL, with HALT showing the greatest AL control. While the retrospective design warrants cautious interpretation, results support the use of all three lenses. Future prospective head-to-head studies are needed to clarify durability and optimal combination strategies. Financial Disclosures:The authors have no proprietary or commercial interest in any materials discussed in this article.
PURPOSE:Fusarium spp. keratitis causes significant corneal blindness and often responds poorly to antifungal therapy. Understanding its pathogenesis is critical for improving disease management. In this context, dysregulated microRNAs are emerging as key modulators of host immune responses and corneal damage. This study aimed to characterize the dysregulation of human corneal miRNAs and their cognate targets during Fusarium spp. keratitis and to examine their association with disease progression, severity, and clinical outcome, including their diagnostic and prognostic potential. METHODS:Corneal swabs and tissues were collected at presentation, after antifungal therapy, and post-therapeutic penetrating keratoplasty in non-responders, alongside cadaveric controls. Small RNA sequencing identified differentially expressed miRNAs at each stage. Functional enrichment and network analyses were conducted. Seventeen miRNAs were validated by RT-qPCR in 10 samples per group, followed by receiver operating characteristic (ROC) analysis in 20 additional samples to determine prognostic accuracy. RESULTS:Distinct miRNA dysregulation occurred across Fusarium spp. keratitis stages. Enrichment analyses revealed involvement of inflammatory pathways, including IL-17, TNF, NF-kappaB, and Toll-like receptor signalling, regulated by miRNAs such as miR-149-5p, miR-184, and miR-328-3p. Downregulation of TGF-β signalling and WNT signalling mediated by miR-16-5p and miR-19a-3p was linked to impaired wound healing in non-responders. RT-qPCR confirmed significant expression changes with strong concordance to sequencing data. ROC analysis identified miR-19a-3p and miR-16-5p as robust prognostic markers across disease stages. CONCLUSION:This study reveals miRNA-mediated mechanisms contributing to Fusarium spp. keratitis pathogenesis. miR-19a-3p and miR-16-5p are promising prognostic biomarkers for severity and treatment response, supporting their relevance for future clinical stratification studies.
PURPOSE:To develop and validate an artificial intelligence (AI)-enabled eye rubbing detection tool using sensor data collected from wrist-based wearable devices. METHODS:An automated system was designed to detect eye rubbing using wrist-based wearable devices. The system involves 3 components: sensor data acquisition, data preprocessing, and deep learning-based classification model. Six different deep learning architectures were developed, including 1D and 2D CNN-LSTM models and an ensemble, to identify the most effective approach. Two datasets were established in the time and frequency domains: a timeseries dataset contains 8640 recordings and a scalogram dataset 15 comprising 112,320 images from 20 subjects. RESULTS:The proposed system demonstrated strong performance, achieving an F1-score of 95.27 ± 0.87% and AUC of 98.26 ± 0.92% across 5 cross-validation folds when using the 1D CNN-LSTM model to distinguish eye rubbing and noneye rubbing activities. When evaluated on the testing set, the system maintained high performance, with an F1-score of 92.54% and AUC of 96.70%. Model inference required 15.32 milliseconds per segment, supporting real-time operation and practical deployments. CONCLUSIONS:The proposed system provides high reliability in detecting eye rubbing behaviors, indicating its potential as a tool to support ophthalmologists and researchers in the rigorous study of the contributions of eye rubbing to the development and progression of keratoconus and other corneal ectasias.
