Search Results (539)

Diabetic retinopathy (DR) reflects retinal microvascular injury and systemic immune-metabolic stress, and most public DR transcriptomic datasets lack paired microbiome/metabolomic profiles. We used gutMGene v2.0 as a curated microbe/metabolite–host gene prior and integrated it with peripheral blood transcriptomics from GSE221521. Candidate genes were refined by weighted gene co-expression network analysis (WGCNA), repeated resampling, cross-dataset assessment, mechanism scoring, peripheral blood mononuclear cell (PBMC) single-cell localization and filamin A (FLNA)-centered single-cell gene regulatory network (GRN) virtual knockout. The gutMGene prior contained 238 host genes; 15 DR-associated genes overlapped this prior, and WGCNA retained ten candidate gut microbe and microbial metabolite-related genes (GMMRGs): FLNA, AKT1, IRAK1, BCL10, CDK6, CTSD, JUP, CXCL1, CXCR2 and IL4R. Resampling prioritized FLNA as the most consistent candidate. Cross-dataset assessment localized the strongest signal to type 2 diabetes (T2D) PBMCs, retinal endothelial cells and advanced proliferative diabetic retinopathy with diabetic macular edema (PDR + DME) retinal tissue, with weaker separation in whole blood, broad retinal tissue and six-donor type 1 diabetes (T1D) PBMCs. FLNA virtual knockout predicted cell-context-dependent perturbation of immune-related transcriptional programs, including IL4R in DR B cells and CTSD in DR monocytes/NK cells. This prior-guided study identifies FLNA within a ten-gene GMMRG set as a circulating host-response signal that links curated microbe/metabolite–host records to immune-vascular and cytoskeletal remodeling in DR. Full article

Background/Objectives: Faricimab is a novel bispecific antibody simultaneously inhibiting vascular endothelial growth factor-A (VEGF-A) and angiopoietin-2 (Ang-2). Its efficacy in vitrectomized eyes with diabetic macular edema (DME), a setting with altered intravitreal pharmacokinetics, remains poorly characterized. We evaluated the 12-month outcomes of intravitreal faricimab (IVF) for DME in eyes vitrectomized for proliferative diabetic retinopathy (PDR). Methods: We retrospectively reviewed 12 consecutive eyes of 11 patients (mean age 59.4 ± 10.5 years) with DME after pars plana vitrectomy (PPV) for PDR who received IVF (6 mg/0.05 mL) between September 2022 and August 2024 with at least 12 months of follow-up. Best-corrected visual acuity (BCVA, logMAR) and central retinal thickness (CRT) were assessed at baseline (BL), 6 months (M6), and 12 months (M12). Results: Eight eyes were treatment-naïve and 4 had been switched from prior anti-VEGF therapy. Two eyes (16.7%) required a change in therapy: one for intraocular inflammation and one for inadequate anatomical response. In the 10 eyes that continued IVF, the mean number of injections was 4.1 ± 1.9. CRT decreased significantly from 489.5 ± 95.9 µm at BL to 328.5 ± 74.7 µm at M6 (p = 0.0065) and 307.0 ± 64.3 µm at M12 (p = 0.0023; repeated-measures ANOVA with Dunnett’s post hoc test). Mean BCVA improved from 0.33 ± 0.26 to 0.23 ± 0.21 logMAR at M12 (p = 0.0894). Conclusions: In this small retrospective study of vitrectomized eyes with DME after PPV for PDR, IVF was associated with significant anatomical improvement, while visual acuity remained stable over 12 months, with a relatively low injection burden. Faricimab may be a useful therapeutic option in this challenging population, although larger prospective studies are warranted to confirm these findings. Full article

Background and Objectives: Faricimab, the first bispecific antibody targeting VEGF-A and angiopoietin-2, has demonstrated durable efficacy in pivotal phase 3 trials for neovascular age-related macular degeneration (nAMD) and diabetic macular edema (DME). Real-world data on treatment-naïve patients managed with fixed-interval maintenance protocols, particularly from South-Eastern Europe, remain limited. This pilot study evaluated 24-month outcomes of intravitreal faricimab in treatment-naïve nAMD and DME, using a standardized four-injection loading phase followed by fixed every-16-week (Q16W) maintenance. Materials and Methods: This study conducted a retrospective, observational, single-center pilot cohort study of 20 consecutive treatment-naïve eyes (9 nAMD, 11 DME). All patients received four monthly loading injections followed by a fixed every-16-week (Q16W) maintenance schedule, supplemented by discretionary additional injections for residual or recurrent disease activity (215 injections total; mean 10.75 ± 0.79 per patient; range 9–12). Primary outcomes were changes in central foveal thickness (CFT) and best-corrected visual acuity (BCVA; Snellen lines with ETDRS letter equivalents) at months 4 and 24. Prespecified secondary analyses included bootstrap 95% confidence intervals, a linear mixed-effects model with a time × disease-group interaction, Bayesian credible intervals with weakly informative priors, false-discovery-rate (FDR) correction, and a minimum detectable effect-size analysis. Results: All 20 eyes completed 24-month follow-up. In nAMD, mean CFT decreased by 186.9 ± 71.9 µm (35.9%; bootstrap 95% CI 148.1–236.0; p < 0.001; d = 2.60), and BCVA improved by 3.89 ± 0.78 Snellen lines (~19 ETDRS letters; 95% CI 3.44–4.33; p < 0.001; d = 4.97). In DME, CFT decreased by 197.7 ± 65.7 µm (39.3%; 95% CI 162.5–237.3; p < 0.001; d = 3.01), and BCVA improved by 4.55 ± 1.04 lines (~23 ETDRS letters; 95% CI 4.00–5.09; p < 0.001; d = 4.39). All 20 eyes (100%) achieved ≥ 3 Snellen lines gain and ≥20% CFT reduction; 80% reached final BCVA ≥ 7 lines. A linear mixed-effects model showed a significant time effect (p < 0.001) but no time × group interaction (CFT p = 0.84; BCVA p = 0.51), indicating concordant trajectories across diseases. Bayesian analysis with weakly informative priors yielded posterior P(|d| > 0.8) ≥ 0.99 for all primary outcomes. After FDR correction, all pre-specified primary comparisons remained significant. The minimum detectable effect size with the realized sample sizes (Cohen’s d ≈ 0.66 combined, 1.07 nAMD, 0.94 DME at 80% power) was substantially below all observed effect sizes. No ocular or systemic adverse events were recorded. Conclusions: In this small, single-center, treatment-naïve pilot cohort, a fixed Q16W faricimab maintenance schedule with discretionary additional injections was associated with durable anatomical and functional improvements over 24 months in both nAMD and DME, with no adverse events recorded across 215 injections. Given the limited sample, these findings should be regarded as hypothesis-generating. The high responder rates likely reflect the cohort’s substantial baseline visual impairment (mean baseline BCVA ~20/120–20/200), which provides greater absolute capacity for measurable gain than in higher-acuity registration trial populations. These pilot data support fixed-interval faricimab as a logistically feasible candidate strategy in resource-constrained settings and should be confirmed in larger multicenter cohorts using standardized ETDRS acuity assessment. Full article

Objective: To predict pixel-wise retinal thickness maps from near-infrared (NIR) fundus images using deep learning (DL), and to identify image features in NIR fundus photographs serving as surrogate markers of retinal thickness, with implications for decentralized diabetic macular edema (DME) screening, progression monitoring, and treatment assessment. Methods: A DL model based on a U-Net architecture was trained on paired NIR fundus and OCT images from 531 eyes across three groups: healthy controls, diabetic retinopathy (DR) without DME, and DME. Model performance was evaluated using mean absolute error (MAE), root mean squared error (RMSE), structural similarity index (SSIM), and center-involved DME (ci-DME) classification at a central subfield thickness threshold of 300 µm. Controlled image manipulation experiments, including spatial disruption of vascular patterns, relocation of hard exudates, and contrast enhancement, were performed to identify image-level features serving as surrogate markers of retinal thickness. Results: The model achieved an MAE of 30.41 ± 18.68 µm, RMSE of 36.14 ± 21.05 µm, and SSIM of 0.87 ± 0.04 across the macula, with consistent performance across ETDRS subfields. For ci-DME classification, it achieved an accuracy of 84.1%, sensitivity of 69.1%, and specificity of 88.7%. Interpretability analyses were performed as qualitative assessments to visualize image regions contributing to model predictions. These analyses highlighted retinal vascular structures, hard exudates, and local contrast variations as visual features observed in relation to model outputs. Conclusions: NIR fundus images contain sufficient structural information to support pixel-wise retinal thickness estimation, with vascular architecture, hard exudates, and local contrast variations identified as image features potentially associated with model predictions. These findings suggest that NIR-based deep learning approaches may have potential applications in the assessment of diabetic macular edema and warrant further prospective and external validation to determine their role in screening, triage support, longitudinal monitoring, and treatment-related assessment, particularly in decentralized and re-source-limited care environments. Full article

Background: Brolucizumab has demonstrated efficacy for diabetic macular edema (DME) in pivotal clinical trials; however, comparative real-world data directly contrasting treatment-naive and anti-VEGF-switched eyes over a prolonged follow-up period remain limited. This case series evaluated 48-week functional, anatomical, and safety outcomes of intravitreal brolucizumab (IVBr) in treatment-naive and switched DME eyes in routine clinical practice. Methods: This retrospective, two-center case series included 21 eyes with center-involving DME treated with IVBr between May 2022 and April 2024. Eyes were classified into a treatment-naive group (n = 10), which included only eyes with no prior anti-VEGF treatment for DME, and a switched group (n = 11), which included eyes previously treated with other anti-VEGF agents but not achieving dry macula. Dry macula was defined as the absence of fovea-involving fluid on OCT. BCVA (logMAR), central retinal thickness (CRT), dry macula rate, the proportion of eyes gaining ≥2 lines of BCVA, number of injections, mean injection interval, and intraocular inflammation (IOI) were assessed at baseline and at 6, 12, 24, and 48 weeks. Results: At baseline, BCVA was 0.44 ± 0.27 in the treatment-naive group and 0.46 ± 0.28 in the switched group (p = 0.943), and CRT was 435.8 ± 150.1 μm and 461.8 ± 139.8 μm, respectively (p = 0.751). In the treatment-naive group, BCVA improved from 0.44 ± 0.27 to 0.24 ± 0.19 at week 48 (p = 0.023), and 5 eyes (50.0%) gained ≥2 lines. In the switched group, BCVA did not improve significantly (0.46 ± 0.28 to 0.41 ± 0.40; p = 0.461); 3 eyes (27.3%) gained ≥2 lines, and 1 eye (9.1%) lost ≥2 lines. CRT decreased significantly in treatment-naive eyes (435.8 ± 150.1 to 312.6 ± 106.2 μm, p = 0.020) and showed a non-significant reduction in switched eyes (461.8 ± 139.8 to 363.6 ± 127.4 μm, p = 0.067). Dry macula rates at week 48 were 40.0% and 27.3%, respectively (p = 0.659). Six switched eyes had prior intravitreal corticosteroid treatment for DME. One treatment-naive eye developed IOI, which resolved without permanent visual impairment. Conclusions: In this retrospective case series, IVBr was associated with anatomical improvement over 48 weeks, with clearer functional and anatomical responses in treatment-naive eyes than in switched eyes. However, the switched group was older and had a more heterogeneous treatment history, including prior intravitreal corticosteroid treatment. The findings should therefore be interpreted as exploratory. Full article

Background: Faricimab is one of the most potent anti-vascular endothelial growth factors used in the management of diabetic macular edema (DME). However, real-world benefits regarding its timing and efficacy are still being explored. Methods: This retrospective non-randomized pilot study aimed to evaluate the efficacy of intravitreal faricimab in the treatment of DME. Eyes initially treatment-naïve for DME with a follow-up of 1 year were grouped as: group 1, where faricimab was introduced within the first six months after the start of treatment; group 2, where it was initiated six or more months after treatment with other drugs. Study parameters included changes in best corrected visual acuity (BCVA) and optical coherence tomography based structural parameters within the 6 × 6 mm optical coherence tomography (OCT) scan regions. Results: Forty-two eyes from 26 patients were analyzed. No statistically significant differences were observed between the groups in cluster-weighted proportions of intra- or sub-retinal fluid, retinal thickness or volume parameters, although group 1 showed modest numerical benefits. SRF showed a trend towards qualitative reduction in group 1, although IRF showed persistence in both groups. Adjusted linear mixed-effects modeling demonstrated no significant impact of early faricimab initiation on functional and anatomical outcomes, which appeared to be influenced by the baseline BCVA, glycemic control, and the number of injections, nullifying the benefits. Conclusions: Faricimab demonstrated modest anatomical improvements with earlier treatment in eyes initially treatment-naïve for DME. Further prospective studies are indicated to assess the treatment strategy and the timing of introduction with faricimab in such eyes. Full article

Background/Objectives: Diabetic retinopathy (DR) is a leading cause of preventable visual impairment, and systematic screening is essential to detect sight-threatening stages before symptoms occur. However, long-term real-world evidence on visual outcomes from structured screening programmes remains limited. This study evaluates, in a retrospective analysis of routinely collected clinical data, the real-life visual outcomes of screening for and treating sight-threatening DR, based on 32 years of data collected in a purpose-built centre implementing the 1990 European Working Party recommendations for screening. Methods: Screening was performed by retinal photography between 1991 and 2022 in 18,161 patients (63,289 screening episodes). Diabetes specialists graded photographs, referring patients to ophthalmologists when needed. The 10-year trajectories of visual acuity (VA) were assessed in patients with different stages of retinopathy and macular involvement at first screening. Results: At first screening, two-thirds of patients had no DR, 15% had mild DR, and the remainder had referable DR or a non-assessable fundus. There was no difference by sex. Patients with more severe DR at first screening had lower VA, but this did not worsen over 10 years. Median VA in 514 patients treated with panretinal photocoagulation for pre-proliferative or proliferative DR changed from 0.10 logMAR (7–8/10) to 0.15 (6–8/10), 1874 ± 1252 days after treatment. In 823 patients photocoagulated for diabetic macular edema, median VA remained 0.10 logMAR (7–8/10) before treatment and 1998 ± 1288 days after treatment. Conclusions: Screening for sight-threatening DR using European Working Party recommendations was feasible in everyday practice and was associated with long-term preservation of visual acuity and a low incidence of severe visual loss. Full article

Objective: This study aimed to evaluate the predictability of baseline optical coherence tomography angiography (OCTA) metrics utilizing a specialized capillary–large vessel segmentation analysis framework in patients with diabetic macular edema (DME) undergoing anti-vascular endothelial growth factor (anti-VEGF) therapy. Methods: Forty-two treatment-naïve eyes with DME receiving three monthly loading anti-VEGF injections were included. Superficial capillary plexus (SCP) images from 3 × 3 mm OCTA scans were processed to isolate the capillary network from the large vessels via image processing. Vessel density and skeleton density were extracted for the total, large-vessel, and capillary components. Multiple linear and logistic regression models were used to identify independent predictors of post-treatment best-corrected visual acuity (BCVA) and “good visual outcome” (≥3-line improvement or final BCVA of 20/40 or better). Results: Following three monthly anti-VEGF injections, the mean BCVA significantly improved from 0.57 ± 0.36 to 0.37 ± 0.30 LogMAR (p < 0.0001), and the mean central retinal thickness decreased from 424.3 ± 117.7 μm to 316.9 ± 84.7 μm (p < 0.0001). The proportion of patients who achieved a good visual outcome was 73.8%. Baseline central retinal thickness was associated with baseline BCVA (p = 0.049) but not predictive of post-treatment BCVA (p = 0.38) or good visual outcomes (p = 0.79). Baseline capillary vessel density was identified as a significant independent predictor of post-treatment BCVA (p = 0.024), whereas total and large-vessel metrics were not. Capillary vessel density was also the only significant predictor of good visual outcomes (p = 0.044). Conclusions: Baseline capillary vessel density is a robust predictor of visual prognosis after anti-VEGF therapy in patients with DME, underscoring the importance of capillary network integrity in functional recovery. Full article

(1) Background: Ultra-early optical coherence tomography (OCT) changes following intravitreal injection may reflect transient mechanical compression rather than pharmacologic effects; however, this temporal profile has not been rigorously characterised with appropriate statistical methodology. (2) Methods: In this prospective observational study, 40 eyes of 40 consecutive patients (one per patient) with macular edema secondary to neovascular age-related macular degeneration (nAMD), diabetic macular edema (DME), or chronic central serous retinopathy (CSR) underwent intravitreal bevacizumab (n = 35) or triamcinolone acetonide (n = 5). Goldmann applanation tonometry and spectral-domain OCT were performed at baseline, 2–5 min, 15 ± 5 min, 24 h, and 48 h post-injection. Repeated-measures ANOVA with Greenhouse–Geisser correction, linear regression, and Spearman rank correlation were applied. (3) Results: Central subfield thickness (CST) decreased markedly at 15 ± 5 min (mean −24.8 ± 11.5%; 95% CI: −28.5% to −21.1%; p < 0.001; partial η² = 0.70), with near-complete rebound by 48 h (−1.0%; p = 0.400). Peak intraocular pressure (IOP) elevation correlated with CST reduction (Spearman r_s = 0.61; 95% CI: 0.39–0.77; p < 0.001), and baseline CST predicted thinning magnitude (R² = 0.52; p < 0.001). (4) Conclusions: Ultra-early OCT thinning after intravitreal injection is consistent with transient mechanical compression. Retinal thickness measurements within 48 h post-injection should be interpreted with caution when assessing treatment response, as early anatomic reduction may not reflect pharmacologic efficacy. Full article

Background: Intravitreal injections (IVI) are widely used for the management of retinal diseases, yet younger adults are underrepresented in clinical trials and real-world reports. Data on the etiologic distribution and treatment patterns of IVI in patients aged ≤50 years remain limited. This study aimed to characterize the indications, treatment strategies, and visual outcomes of IVI in this age group. Material and Methods: Retrospective, single-center observational study including adults aged 18–50 years who received IVI between January 2020 and December 2023 at a tertiary referral hospital in Spain. Demographic data, diagnosis, treatment modality, regimen, number of injections and best-corrected visual acuity (BCVA) were collected. One eye per patient was included for analysis. Visual outcomes were assessed as change in logMAR BCVA between baseline and final follow-up. Results: A total of 122 patients were included. The most frequent indications were diabetic macular edema (29.5%), macular neovascularization of various etiologies (21.3%), retinal vein occlusion (13.9%) and uveitis (9.8%). Anti-vascular endothelial growth factor (anti-VEGF) agents were used in 80.3% of eyes, corticosteroids in 6.6% and combination therapy in 13.1%. The mean number of injections per patient was 6.0 ± 5.3 over a mean follow-up of 3.0 ± 2.5 years. Overall BCVA improved significantly from 0.50 ± 0.59 to 0.38 ± 0.50 logMAR (p = 0.012). Conclusions: IVI in adults ≤ 50 years is uncommon but encompasses a broad etiologic spectrum. Diabetic macular edema and macular neovascularization of diverse etiologies were the leading indications. Anti-VEGF therapy represented the main treatment modality in this cohort. Full article

Uveitis is a heterogeneous group of intraocular inflammatory diseases and an important cause of visual impairment worldwide. Although current treatments mainly target inflammation, many patients develop chronic or recurrent disease, suggesting that inflammation control alone may not fully restore intraocular homeostasis. Increasing evidence highlights the blood–ocular barrier (BOB), including the blood–retinal barrier and blood–aqueous barrier, as a key regulator of the intraocular microenvironment. This review aims to summarize the bidirectional interaction between intraocular inflammation and blood–ocular barrier dysfunction in uveitis, and to highlight the clinical significance of barrier dysfunction in disease monitoring and management. In addition, this review discusses the potential value of incorporating barrier assessment into dynamic disease evaluation and relapse-aware management strategies. Recent studies suggest that inflammation and BOB dysfunction are bidirectionally linked. Inflammatory mediators disrupt barrier integrity, while barrier breakdown facilitates immune cell infiltration and further amplifies inflammation, forming a self-reinforcing cycle that may drive disease persistence. Importantly, BOB dysfunction also has clinical implications. Findings such as aqueous flare, macular edema on optical coherence tomography, and vascular leakage on fluorescein angiography reflect barrier status and can serve as dynamic indicators for disease monitoring. Persistent abnormalities despite reduced inflammatory cell activity may indicate incomplete barrier recovery or subclinical inflammation, helping to explain discordant clinical findings and the relapse-prone nature of uveitis. Rather than viewing BOB dysfunction solely as a pathological consequence of inflammation, this review highlights the potential clinical value in disease assessment and management. Barrier-related findings may provide additional information beyond conventional inflammatory evaluation, particularly in cases where inflammatory cell activity appears controlled, but underlying barrier alteration persists. Incorporating barrier assessment into monitoring may help interpret discordant clinical findings, improve evaluation of disease control, and support a more relapse-aware management strategy in uveitis. In addition, therapeutic approaches aimed at restoring barrier integrity may provide a more comprehensive strategy for achieving sustained remission and reducing recurrence risk. Full article

Background/Objectives: Retinal layer segmentation in optical coherence tomography (OCT) is essential for the diagnosis and monitoring of retinal diseases such as age-related macular degeneration (AMD) and diabetic macular edema (DME). Although deep learning methods have achieved strong performance, most rely on dense pixel-wise predictions and often struggle to preserve anatomical consistency, particularly in regions with low contrast or structural deformation. This study aims to address these limitations by introducing a query-based segmentation framework that explicitly models retinal layer structure. Methods: In this paper, we propose the RetiQueryNet architecture that employs encoding of retinal layers in the form of query embeddings with the use of cross attention to interact with pixel level features encoded by a transformer based encoder. The architecture integrates multi-scale features through a compact query-driven decoder with modest additional computational overhead. Normalization and resizing of OCT images preceded their usage as inputs, while the layer labels were converted to multi-class segmentation maps. In the training process, we used loss function with combination of cross entropy loss and Dice loss. Our model performance was compared with multiple state-of-the-art models such as U-Net, DeepLabV3, FPN, MANet and SegFormer, while performance metrics were Dice, IoU and mean surface distance (MSD). Results: RetiQueryNet was able to attain a mean Dice score of 0.934 ± 0.0046 and outperformed all baseline models on the main performance measures. Improvements were particularly evident in challenging retinal layers such as IBRPE and OBRPE, where boundary ambiguity is high. It should be noted that RetiQueryNet had a relatively lower MSD value, meaning that the predicted boundaries were more accurate. Furthermore, visual observations suggest that the approach generated smooth and coherent segmentations. Conclusions: The findings demonstrate that query-based modeling offers a viable approach to pixel-wise segmentation. In particular, by making use of structural priors in the form of learnable queries, RetiQueryNet improves not only segmentation accuracy but also anatomical consistency. Query-based modeling appears to be an exciting area for retinal image segmentation that could potentially be applied to other applications in medical image segmentation. Full article

Background/objective: Suprachoroidal triamcinolone acetonide (TA) was recently FDA-approved and is emerging as a new alternative to other local therapies for macular edema (ME) associated with noninfectious uveitis (NIU). The objective of this study is to evaluate the preliminary safety and efficacy of suprachoroidal TA in patients with refractory ME secondary to NIU. Methods: This was a retrospective review of a small cohort of patients with refractory ME secondary to NIU treated with suprachoroidal TA from November 2022 to October 2023. Results: Six eyes from five patients with refractory ME secondary to NIU were included in the study. The cohort included two females (40%), and the median age was 62 years (IQR = 8). Ophthalmic diagnoses included intermediate uveitis (n = 2; 40%), birdshot chorioretinopathy (n = 1; 20%), autoimmune retinopathy (n = 1; 20%), and panuveitis (n = 1; 20%). The median logMAR visual acuity was 0.7 (Snellen 20/100) at baseline and improved to 0.3 (Snellen 20/40) during follow-up visits at 1 month and 2–3 months. The median central subfield thickness (CST) was 690 μm at baseline and improved to 367.5 μm and 309 μm at the follow-up visits at 1 month and 2–3 months, respectively. The initial improvement in logMAR visual acuity and CST was less pronounced at follow-up visits at 6–7 months and 11–12 months. Conclusions: This study demonstrates the safety of suprachoroidal TA and efficacy signals, including improvement in visual acuity and ME at 3 months in patients with severe, refractory ME secondary to NIU. Full article

Background/Objectives: Disorganization of retinal inner layers (DRIL) is an important and supportive biomarker in optical coherence tomography (OCT) imaging for diagnosing the extent of diabetic macular edema (DME) in patients and anticipating visual outcomes. But the manual DRIL identification is subject to interobserver bias and requires a lot of time and effort from the experts. This research presents a novel, computerized, and clinically guided approach for the classification of DRIL that leverages the central 1 mm foveal region extracted through the annotations provided by the expert ophthalmologists and investigates the effectiveness of a transformer and Masked Auto Encoder (MAE) based foundation model (RETFound) as the primary approach. Methods: We fine-tuned and validated the RETFound model, utilizing accurate foveal center coordinates provided by the experienced ophthalmologists. Our approach emphasizes the macular region that is significant diagnostically, where DME biomarkers manifest more predominantly. To guarantee robust evaluation, the dataset was divided into 85% training and 15% held-out test sets. We performed 5-fold cross-validation exclusively on the training dataset with baseline, conservative, and moderate fine-tuning strategies, and the final model was evaluated on the independent, unseen test set. Convolutional neural network (CNN)-based transfer learning (TL) models (MobileNetV2, EfficientNetB0, InceptionV3, DenseNet121, and DenseNet169) were also assessed for comparative evaluation. Results: The RETFound model yielded the best outcomes under the conservative fine-tuning strategy, achieving a mean test accuracy (AC) of 0.9339 ± 0.0036 and an area under the curve (AUC) of 0.9660 ± 0.0028 on the independent held-out test set across the five fold-trained models. The moderate and baseline evaluations achieved comparatively lower outcomes, highlighting the effectiveness of the conservative approach. The RETFound model consistently outperformed CNN models, exhibiting stability and superior generalization for DRIL classification. We performed statistical validation using the Wilcoxon signed-rank test and 95% confidence intervals to confirm the robustness of the proposed method, and an ablation analysis showed that the fovea-centered region of interest (ROI) guidance consistently improved results when compared with whole OCT analysis. Conclusions: This research demonstrates that the deep-learning (DL) methods assisted by expert clinical knowledge with an anatomically aligned ROI could provide remarkable results in DRIL detection applications. This work attempts to establish an anatomically relevant framework for computerized DRIL identification that focuses on the highly crucial macular region, possibly helping in faster intervention and improved diagnosis in the management of DME. Full article

Background/Objectives: This retrospective cohort study compared hospitalization and follow-up rates in patients with newly diagnosed proliferative diabetic retinopathy (PDR) versus those without prior diabetic retinopathy (DR) screening. Methods: Using TriNetX, a global electronic health record database, 57,964 patients aged ≥ 40 years with type 2 diabetes and newly diagnosed PDR without diabetic macular edema (DME) requiring panretinal photocoagulation or intravitreal injection were included. Patients were stratified based on the presence or absence of prior DR screening in the last 5 years and balanced using propensity score matching (PSM). Primary outcomes included 30-, 60-, and 90-day hospitalization rates and repeat ophthalmic follow-up as estimated using repeat PDR diagnosis codes and repeat retinal imaging codes, including OCT, fundus photography, and fluorescein angiography. Results: Of 57,964 patients, 25,003 had no prior DR screening and 32,961 had prior DR screening. After matching, 19,316 patients were included per cohort. Patients without known DR screening had significantly higher hospitalization rates at 30 days (RR = 1.78, 95% CI 1.67–1.89), 60 days (RR = 1.59, 95% CI 1.51–1.67), and 90 days (RR = 1.51, 95% CI 1.44–1.58), and lower repeat ophthalmic visits by PDR codes at 30 days (RR = 0.458, 95% CI 0.440–0.476), 60 days (RR = 0.450, 95% CI 0.437–0.463) and 90 days (RR = 0.420, 95% CI 0.408–0.432) or by repeat retinal imaging codes at 30 days (RR = 0.450, 95% CI 0.423–0.478), 60 days (RR = 0.394, 95% CI 0.377–0.411), and 90 days (RR = 0.381, 95% CI 0.366–0.396) (all p < 0.0001). Conclusions: Absence of known prior DR screening in PDR patients is associated with higher hospitalization risk and reduced ophthalmic follow-up, suggesting that a lack of screening indicates broader gaps in healthcare engagement and disease control. Tailored strategies are needed to prevent vision loss as well as systemic complications. Full article