School of Medicine

PURPOSE: To evaluate demographic and socioeconomic factors associated with pediatric ophthalmology referrals from the UC San Diego (UCSD) EyeMobile, a school-based mobile vision program. METHODS: The records of children screened from 2021 to 2025 were reviewed retrospectively. Children who failed a screening received a comprehensive eye examination by an optometrist. Referrals to a pediatric ophthalmologist were made for those with potentially significant ocular pathology and no established eye care provider. Socioeconomic status was assessed using the national Child Opportunity Index (COI), analyzed as continuous scores and quintiles. Referral rates were compared across quintiles using two-proportion z tests, and multivariable logistic regression was used to identify factors associated with referrals. RESULTS: Of 42,166 children screened during the study period, 5,657 (14.1%) received a comprehensive examination, and 217 (0.51%) were referred to an ophthalmologist. Leading reasons for reasons included significant refractive errors (39.2%), strabismus (18.4%), and suspicion for keratoconus (14.3%). Referral rates were 0.69% in Very Low COI areas and 0.34% in Very High COI areas. There was a 7.6% reduced odds of referral per 10-point increase in COI (OR = 0.934, P = 0.033). Compared with White children, Black (OR = 2.30, P = 0.012) and Hispanic children (OR = 2.03, P = 0.011) had significantly higher odds of referral. CONCLUSIONS: Lower socioeconomic status and minority race/ethnicity were independently linked to higher odds of pediatric ophthalmology referral, highlighting the disproportionate disease burden in underserved groups. Mobile screening programs may help with timely identification of at-risk children and improve health equity.

BACKGROUND/AIMS: To apply retinal nerve fibre layer (RNFL) optical texture analysis (ROTA) to investigate (1) the patterns of RNFL bundle defects, and (2) the frequency of papillomacular and papillofoveal bundle involvement across early, moderate and advanced glaucoma. METHODS: All eyes underwent 24-2 visual field (VF) testing and optical coherence tomography (OCT) for ROTA. The borders of RNFL defects were delineated from ROTA, and the involvement of the arcuate, papillomacular and papillofoveal bundles was determined for each eye. 24-2 VF stimulus projections were mapped onto the corresponding topographic areas of ROTA images. Multilevel logistic regression analysis was applied to evaluate the structure-function association. RESULTS: Papillomacular bundle defects were highly prevalent in glaucoma, increasing from 87.7% in early to 95.35% in moderate and 100% in advanced glaucoma. Papillofoveal bundle defects were also common, increasing from 29.7% in early to 36.05% in moderate and 60.98% in advanced glaucoma. Central four 24-2 test locations that projected onto the trajectories of papillomacular or papillofoveal RNFL bundle defects demonstrated significantly increased likelihood of VF sensitivity abnormality (ORs of 22.42 at PDP<5% and 20.26 at TDP<5%, respectively, p<0.001 for both). CONCLUSION: ROTA uncovers a wide spectrum of RNFL bundle defects spanning the entire glaucoma continuum. It also provides visualisation of the preserved RNFL bundles in advanced glaucoma. Papillomacular and papillofoveal RNFL bundle defects are present in a considerable proportion of eyes with early, moderate and advanced glaucoma, and, when detected, they significantly increase the likelihood of abnormality in the corresponding central 24-2 test locations.

Objective Local injections such as bupivacaine are used to treat strabismus, although the mechanism of its clinical effects is unknown. The purpose of this study is to analyze the effects of bupivacaine on extraocular muscle regeneration in a pre-clinical model. Design Laboratory study. Subjects, Participants, and/or Controls Three-month-old male New Zealand white rabbits with age- and sex-matched saline injection controls and un-injected controls. Methods One superior rectus per animal was injected with either 3% bupivacaine or saline control and harvested at 1 week, 1 month, or 3 months after treatment. Histomorphometric analysis was used to determine the effects of bupivacaine on extraocular muscles. Data were analyzed for the global and orbital layer separately, as well as the entire muscle cross-sectional area. Main Outcome Measures. Muscle morphology, markers of muscle regeneration, muscle size. Results 3% bupivacaine treatment has a myodestructive effect on the injected extraocular muscle with disruption of myofiber organization and areas of missing myofibers. However, even one week after injection, we observed significantly more embryonic myosin heavy chain positive myofibers, which signify newly regenerating myofibers, in the bupivacaine group compared to the saline group. Embryonic myosin heavy chain is still significantly expressed at 1 month after bupivacaine injection in extraocular muscles, in stark contrast to skeletal muscles elsewhere in the body where embryonic myosin heavy chain decreases 1-2 weeks after injury. There is a significant increase in the density of muscle stem cells and myofibers with centralized nuclei 3 months after bupivacaine injection compared to saline in both the orbital layer and global layer. Muscle size and extracellular matrix components did not increase at 3 months after bupivacaine injection compared to saline. Conclusions Bupivacaine’s effects on extraocular muscles may be explained in part by bupivacaine triggering a regenerative response, initially forming new myofibers and increasing the number of muscle stem cells to exert its effects as a treatment for strabismus.

PURPOSE: To evaluate the effects of teprotumumab on individual extraocular muscle (EOM) volume and proptosis in thyroid eye disease (TED). METHODS: A retrospective review was performed using exophthalmometry, clinical activity score (CAS), and orbital imaging before and after teprotumumab treatment. The lateral rectus (LR), medial rectus (MR), superior rectus (SR), inferior rectus (IR) muscles, and the globe were manually segmented individually. Individual and total EOM volumes were calculated and compared between pre- and post-treatment magnetic resonance imaging (MRI). RESULTS: Mean proptosis reduction after teprotumumab treatment was 3.69 mm (95% CI: 2.82, 4.17, p < 0.001), with an improvement in all orbits. Significant muscle volume reductions occurred in the MR (-20%, p < 0.001) and IR (-28%, p = 0.002) but not the LR (-11%, p = 0.063) or SR (-19%, p = 0.053). Total EOM volume decreased by 20%, while globe volume remained unchanged. CAS and proptosis improved post-treatment, but changes in total EOM volume did not significantly correlate with changes in CAS or proptosis. Pre-treatment IR, MR, and SR volumes showed moderate correlation with total EOM volume reduction. CONCLUSIONS: Teprotumumab significantly reduces EOM volumes, CAS, and proptosis in TED. Weak correlations with clinical outcomes suggest that additional factors, such as orbital fat, influence treatment response.

PURPOSE: To determine associations of income and education level with a diagnosis of strabismus and to identify socioeconomic variables that may affect timely access to diagnosis. METHODS: Annual income, highest level of education completed, and ZIP code income, high school completion, poverty, and socioeconomic deprivation metrics were collected from 413,360 participants in the database. A χ² test was used to determine significant differences in distributions of income, education, and ZIP code metrics in 3,734 strabismus patients compared with the overall database. RESULTS: Participants living in ZIP codes with lower multidimensional deprivation indices (less deprivation) are more likely to be diagnosed with strabismus. Participants with annual income below $10,000 (10.10%) or who completed education between fifth grade and a high school diploma or GED (20.06%) are less likely to receive a diagnosis for certain strabismus subtypes. Participants with annual income over $200,000 (7.07%), advanced degrees (27.02%), living in ZIP codes with higher income, and higher high school completion rates were more likely to be diagnosed with certain strabismus subtypes. CONCLUSIONS: Participants diagnosed with strabismus tend to have higher incomes, an advanced degree, and reside in more affluent ZIP codes, whereas those with lower income and education levels are less likely to be diagnosed. These findings reveal potential socioeconomic disparities in access to ophthalmic diagnostic services and care. Findings emphasize the importance of addressing socioeconomic barriers in eye care to enable equitable access.

PURPOSE: To determine the prevalence of adult-diagnosed strabismus and its associations with sex, age, and race in a large, diverse population database in the United States. METHODS: Sex, age, and race data were collected from 413,457 individuals in the All of Us database. The χ² test with post hoc pairwise comparisons was used to determine significant differences in distributions of sex, age, and race data in 3,734 strabismus patients compared with the overall database. RESULTS: There is a higher proportion of males among patients with strabismus compared with the overall database (43.34% vs 38.34% [P < 0.001]). There is a significantly higher proportion of patients aged over 65 years among patients with strabismus compared with the overall database (45.77% vs 24.7% [P < 0.001]). There is also a significantly different racial distribution of individuals with strabismus compared with the overall database (P < 0.001). Pairwise comparisons showed a significantly lower proportion of Asian individuals among patients with strabismus or heterotropia, a lower proportion of Black individuals with most subtypes, and a higher proportion of White individuals among patients with any strabismus. CONCLUSIONS: These results suggest that there is a lower proportion of females, Asian, and Black participants with diagnosed strabismus compared with the overall population in the All of Us database. These findings may indicate that strabismus affects patient populations differently, or that there is differential access to care and diagnosis of strabismus across sex, age, and race.

In this study, we developed deep learning-based method to classify the type of surgery performed for epiretinal membrane (ERM) removal—either internal limiting membrane (ILM) removal or ERM-alone removal. Our model, based on the ResNet18 convolutional neural network (CNN) architecture, utilizes postoperative optical coherence tomography (OCT) center scans as inputs. We evaluated the model using both original scans and scans preprocessed with energy crop and wavelet denoising, achieving 72% accuracy on preprocessed inputs, outperforming the 66% accuracy achieved on original scans. To further improve accuracy, we integrated tunable wavelet units with two key adaptations: Orthogonal Lattice-based Wavelet Units (OrthLatt-UwU) and Perfect Reconstruction Relaxation-based Wavelet Units (PR-Relax-UwU). These units allowed the model to automatically adjust filter coefficients during training and were incorporated into downsampling, stride-two convolution, and pooling layers, enhancing its ability to distinguish between ERM-ILM removal and ERM-alone removal, with OrthLatt-UwU boosting accuracy to 76% and PR-Relax-UwU increasing performance to 78%. Performance comparisons showed that our AI model outperformed a trained human grader, who achieved only 50% accuracy in classifying the removal surgery types from postoperative OCT scans. These findings highlight the potential of CNN based models to improve clinical decision-making by providing more accurate and reliable classifications. To the best of our knowledge, this is the first work to employ tunable wavelets for classifying different types of ERM removal surgery.

This work introduces a novel biorthogonal tunable wavelet unit constructed using a lifting scheme that relaxes both the orthogonality and equal filter length constraints, providing greater flexibility in filter design. The proposed unit enhances convolution, pooling, and downsampling operations, leading to improved image classification and anomaly detection in convolutional neural networks (CNN). When integrated into an 18-layer residual neural network (ResNet-18), the approach improved classification accuracy on CIFAR-10 by $\mathbf{2. 1 2} \boldsymbol{\%}$ and on the Describable Textures Dataset (DTD) by 9.73%, demonstrating its effectiveness in capturing fine-grained details. Similar improvements were observed in ResNet-34. For anomaly detection in the hazelnut category of the MVTec Anomaly Detection dataset, the proposed method achieved competitive and wellbalanced performance in both segmentation and detection tasks, outperforming existing approaches in terms of accuracy and robustness.

Purpose: To review and evaluate the current literature on the application and impact of large language models (LLMs) in the field of ophthalmology, focusing on studies published in high-ranking ophthalmology journals. Design: This is a retrospective review of published articles. Participants: This study did not involve human participation. Methods: Articles published in the first quartile (Q1) of ophthalmology journals on Scimago Journal & Country Rank discussing different LLMs up to June 7, 2024, were reviewed, parsed, and analyzed. Main Outcome Measures: All available articles were parsed and analyzed, which included the article and author characteristics and data regarding the LLM used and its applications, focusing on its use in medical education, clinical assistance, research, and patient education. Results: There were 35 Q1-ranked journals identified, 19 of which contained articles discussing LLMs, with 101 articles eligible for review. One-third were original investigations (32%; 32/101), with an average of 5.3 authors per article. The United States (50.4%; 51/101) was the most represented country, followed by the United Kingdom (25.7%; 26/101) and Canada (16.8%; 17/101). ChatGPT was the most used LLM among the studies, with different versions discussed and compared. Large language model applications were discussed relevant to their implications in medical education, clinical assistance, research, and patient education. Conclusions: The numerous publications on the use of LLM in ophthalmology can provide valuable insights for stakeholders and consumers of these applications. Large language models present significant opportunities for advancement in ophthalmology, particularly in team science, education, clinical assistance, and research. Although LLMs show promise, they also show challenges such as performance inconsistencies, bias, and ethical concerns. The study emphasizes the need for ongoing artificial intelligence improvement, ethical guidelines, and multidisciplinary collaboration. Financial Disclosures: The author(s) have no proprietary or commercial interest in any materials discussed in this article.

Objective: To quantitatively assess the retinal vascular tortuosity of patients with sickle cell disease (SCD) and retinopathy (SCR) using an automated deep learning (DL)-based pipeline. Design: Cross-sectional study. Subjects: Patients diagnosed with SCD and screened for SCR at an academic eye center between January 2015 and November 2022 were identified using electronic health records. Eyes of unaffected matched patients (i.e., no history of SCD, hypertension, diabetes mellitus, or retinal occlusive disorder) served as controls. Methods: For each patient, demographic data, sickle cell diagnosis, types and total number of sickle cell crises, SCD medications used, ocular and systemic comorbidities, and history of intraocular treatment were extracted. A previously published DL algorithm was used to calculate retinal microvascular tortuosity using ultrawidefield pseudocolor fundus imaging among patients with SCD vs. controls. Main Outcome Measures: Cumulative tortuosity index (CTI). Results: Overall, 64 patients (119 eyes) with SCD and 57 age- and race-matched controls (106 eyes) were included. The majority of the patients with SCD were females (65.6%) and of Black or African descent (78.1%), with an average age of 35.1 ± 20.1 years. The mean number of crises per patient was 3.4 ± 5.2, and the patients took 0.7 ± 0.9 medications. The mean CTI for eyes with SCD was higher than controls (1.06 ± vs. 1.03 ± 0.02, P < 0.001). On subgroup analysis, hemoglobin S, hemoglobin C, and HbS/beta-thalassemia variants had significantly higher CTIs compared with controls (1.07 vs. 1.03, P < 0.001), but not with sickle cell trait variant (1.04 vs. 1.03 control, P = .2). Univariable analysis showed a higher CTI in patients diagnosed with proliferative SCR, most significantly among those with sea-fan neovascularization (1.06 ± 0.02 vs. 1.04 ± 0.01, P < 0.001) and those with >3 sickle cell crises (1.07 ± 0.02 vs. 1.05 ± 0.02, P < 0.001). Conclusions: A DL-based metric of cumulative vascular tortuosity associates with and may be a potential biomarker for SCD and SCR disease severity. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.