Search Results (66)

Existing encoder–decoder architectures operating in the field of deep learning-based image segmentation face fundamental limitations such as information loss, performance degradation as network depth increases, and high computational costs. To overcome these issues, we propose a new architecture that integrates features from different depth levels at a single fusion point. This approach enables both comprehensive representation power and the preservation of very small details. The proposed approach creates an efficient structure that achieves high accuracy values without requiring unnecessary network deepening. The designed model was comprehensively compared with state-of-the-art architectures such as U-Net, V-Net, W-Net, T-Net, Seg-Net, and Multiple U-Net, which are accepted in the literature, on datasets with different characteristics such as MedSeg, Retina Drive, and Massachusetts datasets. Experimental findings reveal that the developed method outperforms its competitors in all test metrics. In particular, the dice (DSC) score, the most critical indicator of segmentation accuracy, achieved a value of 0.957 on the Retina DRIVE dataset, demonstrating a significant performance difference compared to existing models that remained in the 0.68–0.81 range in challenging scenarios. Furthermore, the 99.6% accuracy (Acc) and 0.006 loss (Loss) values obtained on COVID-19 CT data confirm the architecture’s error-free learning capacity. The stable loss function trend observed across all datasets demonstrates the model’s stable learning ability and high generalization capability. Full article

Background/Objectives: While CRISPR/Cas9 technology offers a revolutionary approach for correcting genetic ocular blindness, efficient and safe delivery remains the primary bottleneck. Traditional viral vectors, despite their efficacy, face challenges regarding cargo size limitations and potential genomic integration risks. Non-viral vectors offer distinct comparative advantages, including large cargo capacity for diverse CRISPR tools and transient expression to minimize off-target effects, but must overcome the eye’s formidable static and dynamic barriers, specifically the corneal epithelium, vitreous humor, and the inner limiting membrane. In this review, we present an anatomically guided framework for non-viral CRISPR/Cas9 delivery, mapping engineering strategies to specific ocular tissue targets. We first delineate the mechanisms of key physiological barriers, including the corneal stroma, aqueous humor circulation, and the vitreous–retina interface. Subsequently, we critically evaluate the latest advancements in non-viral platforms, such as pH-responsive lipid nanoparticles and engineered virus-like particles. The core focus of this review is on site-specific breakthrough strategies: from utilizing mucoadhesive polymers to counteract tear clearance in the cornea to exploiting specialized administration routes, such as suprachoroidal space and subretinal injection, to bypass retinal barriers, and deep-penetrating intravitreal carriers for targeting the photoreceptor-RPE complex. By integrating material science with precise administration routes, this review highlights feasible translational pathways for next-generation, carrier-free, or biomimetic ocular gene editing therapies. Full article

Background/Objectives: Refractory macular holes (MHs) that persist after conventional internal limiting membrane (ILM) peeling pose a significant surgical challenge. In this study, we analyzed the anatomical and functional outcomes of epiretinal human amniotic membrane (hAM) transplantation in patients with MHs. Methods: This retrospective study included 10 eyes of 10 patients with refractory MHs. All patients underwent 25-gauge pars plana vitrectomy, epiretinal cryopreserved hAM transplantation, and C3F8 gas tamponade. The large hAM graft was placed over the macula with the stromal side facing the retina. Preoperative and postoperative best-corrected visual acuity (BCVA), optical coherence tomography (OCT) findings, and MH dimensions were recorded. Results: The mean follow-up period was 7 months (range: 3–14 months). The mean preoperative minimum linear diameter and base diameter of the MHs were 715 ± 212 μm and 1114 ± 258 μm, respectively. Anatomical closure was achieved in all patients (100%). Postoperative OCT revealed rearrangement of the inner and other retinal layers in 7 out of 10 patients (70%), with partial restoration of the outer retinal layers. The mean logMAR BCVA improved significantly from 1.60 ± 0.37 preoperatively to 1.00 ± 0.45 postoperatively (p < 0.001). No graft dislocation, rejection, or other significant complications were observed. Conclusions: Our preliminary results suggest that epiretinal human amniotic membrane transplantation is a feasible and promising surgical technique for achieving anatomical closure and functional improvement in refractory macular holes in which conventional ILM peeling has failed. Full article

Background/Objectives: Pigment migration is a key biomarker of progression in age-related macular degeneration (AMD). This study assessed the diagnostic performance of ring aperture Retro mode (RAR) imaging for detecting pigment migration and compared its performance with established multimodal imaging techniques. Methods: This retrospective study included 80 eyes from 61 consecutive patients with AMD who underwent multimodal imaging with color fundus images (CFIs), fundus autofluorescence (FAF), RAR imaging (Mirante, NIDEK), and en face optical coherence tomography (OCT) with B-scans (Cirrus HD-OCT 5000, Zeiss). Two independent retina specialists graded the AMD stage and the presence of pigment migration across modalities. Sensitivity and positive predictive value (PPV) of RAR were calculated using en face OCT as the reference standard. Results: RAR demonstrated high diagnostic performance, with a sensitivity of 94.7% and a PPV of 93.4% relative to en face OCT. RAR frequently identified pigment migration that was not visible on CFI or FAF, particularly in early AMD and in eyes with media opacity. Distinct morphologic patterns—including hyperreflective foci, thickened retinal pigment epithelium, refractile drusen, and cuticular drusen—were consistently identifiable on RAR. In four eyes with geographic atrophy, RAR detected perifoveal pigment redistribution at least six months before foveal involvement was confirmed by OCT and FAF. Conclusions: RAR imaging is a rapid, sensitive, and clinically practical technique for detecting pigment migration in AMD. By complementing en face OCT and enhancing visualization in cases where standard imaging is limited, RAR may strengthen early disease surveillance, support prognostic assessment, and improve multimodal diagnostic workflows in routine practice. Full article

Objectives: To evaluate the utility of widefield montage swept-source OCT angiography (SS-OCTA) in detecting and monitoring retinal vasculitis beyond the posterior pole. Methods: Prospective case series. Patients with clinically diagnosed retinal vasculitis imaged with a same-day widefield SS-OCTA montage and ultra-widefield fluorescein angiography (FA) at 2 or more visits. Five overlapping 12 × 12 mm SS-OCTA scans were acquired to provide imaging of the posterior pole and each quadrant of the near periphery. A color retinal thickness map was superimposed on each 12 × 12 mm en-face flow scan with a customized segmentation to demonstrate perivascular retinal thickening. A composite “montage” image was then created by combining the scans to allow for analysis of the macula and near periphery. Findings were then correlated with the same-day FA, the current “gold standard” diagnostic tool for retinal vasculitis, to assess diagnostic efficacy. Results: SS-OCTA demonstrated perivascular thickening in both the posterior pole and peripheral retina in 30 eyes of 16 patients and was found to be an effective diagnostic tool with good correlation to findings on fluorescein angiography for monitoring retinal vasculitis over time. Conclusions: The widefield SS-OCTA montage expands the visualization of retinal vasculitis into the near periphery, providing a noninvasive tool that may complement FA in the diagnosis and monitoring of retinal vasculitis. Full article

Objectives: We propose a flow-signal-based classification of retinal artery macroaneurysms (RAMs) using Optical Coherence Tomography Angiography (OCTA) and compare the findings with fundus fluorescein angiography (FFA). Methods: A retrospective review of 49 RAM cases observed over 6 years (October 2017–March 2023) at a medical retina clinic at the University Hospital Southampton, UK. Electronic clinical records, FFA, and OCTA images (en face and B-scan) were reviewed to identify pathology and assess RAM flow profiles. Results: In total, 30 eyes from 30 patients were included. The mean age of the patients was 76 years (range 49–91), with 17 females and 13 males. All eyes underwent OCTA, enabling classification of RAMs into three flow signal types: high (9 eyes), low (10 eyes), and absent (9 eyes), while 2 eyes had haemorrhage-related artefacts. A subgroup of 13 eyes also underwent FFA, allowing direct comparison, which showed flow profiles similar to those of OCTA: high (4 eyes), low (6 eyes), and absent (2 eyes), with 1 ungradable case due to subretinal haemorrhage masking. A discrepancy in flow was observed in one case where FFA indicated flow, but OCTA did not. Despite this, FFA and OCTA generally agreed on the flow levels, with a Spearman correlation of r = 0.79 (p = 0.004). Conclusions: OCTA flow profiles were directly comparable to FFA. OCTA effectively identified different levels of blood flow signal behaviour in RAMs. The proposed flow-based RAM classification may aid in prognosis, treatment indications, follow-up, and safe repeat imaging in clinical practice without systemic risk to the patient. Full article

Background: Sturge–Weber syndrome (SWS) typically presents with a port-wine stain on the face, accompanied by leptomeningeal capillary malformations and ocular vascular anomalies. The aim of our study was to evaluate retinal vascular density and vessel diameter to better characterize the presence of vascular alterations. Methods: 15 patients diagnosed with Sturge–Weber syndrome and 15 healthy controls underwent comprehensive ophthalmologic evaluation, Optical Coherence Tomography (OCT) and Optical Coherence Tomography Angiography (OCTA), to evaluate the microvascular architecture of the retina and choroid. Results: Analysis of the processed images revealed a significant increase (p < 0.05 *) in the density of the deep capillary plexus in patients with SWS compared to healthy controls. Vascular diameter was found to be increased overall in several retinal vascular plexuses in SWS patients compared to controls, reaching statistical significance (p < 0.05 *) in the deep vascular complex. Conclusions: The findings from our analysis highlight the potential role of OCTA in predicting the progression or worsening of ocular diseases over time. The introduction of new assessment parameters offers additional perspectives in evaluating ocular health. Since this examination allows for the detection of changes in the morphology and density of blood vessels as revealed by OCTA, these changes can be correlated with disease progression and the effectiveness of therapies. Full article

Inherited retinal dystrophies (IRDs) are clinically and genetically heterogeneous disorders. Most IRDs follow a monogenic inheritance pattern. However, an increasing number of unresolved cases suggest the possible contribution of oligogenic or digenic mechanisms. Here, we report two ultra-rare missense variants—AIPL1 R302L and BBS2 P134R—that co-segregate with early-onset nonsyndromic retinal degeneration in affected individuals from a non-consanguineous family. We performed a multi-level computational investigation to assess whether these variants may act through a convergent pathogenic mechanism. Using AlphaFold2-predicted structures, we modeled both wild-type and mutant proteins, introduced point mutations, and performed energy minimization and validation. FoldX, DynaMut2, and DUET all predicted destabilizing effects at the variant sites, corroborated by local disruption of secondary structure and altered surface electrostatics. Comparative docking (via HDOCK and ClusPro) identified a putative interaction interface between the TPR domain of AIPL1 and the β-sheet face of BBS2. This interface was destabilized in the double-mutant model. At the systems level, transcriptomic profiling confirmed co-expression of AIPL1 and BBS2 in human retina and fetal eye, while functional enrichment analysis highlighted overlapping involvement in ciliary and proteostasis pathways. Network propagation suggested that the two proteins may converge on shared interactors relevant to photoreceptor maintenance. Collectively, these in silico results provide structural and systems-level support for a candidate digenic mechanism involving AIPL1 and BBS2. While experimental validation remains necessary, our study proposes a testable mechanistic hypothesis and underscores the value of computational approaches in uncovering complex genetic contributions to IRDs. Full article

Grassland ecosystems are vital for global biodiversity, yet traditional wildlife monitoring methods are often labor-intensive and costly. Although drone-based aerial surveys provide a scalable alternative, they face significant challenges such as detecting extremely small targets, handling complex backgrounds, and operating under strict computational constraints. To address these issues, this study proposes IECA-YOLOv7, a lightweight detection model that incorporates three key innovations: an Improved Efficient Channel Attention (IECA) module for enhanced feature representation, a content-aware CARAFE upsampling operator for improved detail recovery, and a Normalized Wasserstein Distance (NWD) loss function for robust small-target localization. Evaluated on a dedicated grassland wildlife dataset (GWAID), the model achieves a mAP@0.5 of 86.6% and a mAP@0.5:0.95 of 47.2%, outperforming the baseline YOLOv7-tiny by 2.9% in Precision and 1.8% in Recall. Furthermore, it surpasses non-YOLO architectures such as RetinaNet, EfficientDet-D0, and DETR by significant margins, demonstrating superior performance in small-object detection under complex conditions. Cross-dataset validation on VisDrone, CARPK, and DOTA demonstrates a strong generalization capability. With a model size under 5 MB, IECA-YOLOv7 effectively balances accuracy and efficiency, offering a practical solution for real-time wildlife monitoring via drones under challenging environmental constraints such as variable lighting, occlusion, and limited computational resources, thereby supporting broader conservation efforts. Full article

Background: In clinical practice, visual outcomes with anti-VEGF therapy may be worse than those observed in clinical trials. In this study, we aim to investigate the long-term outcomes of neovascularization treated with intravitreal aflibercept injections (IAI) in a teaching hospital setting. Methods: This is a retrospective, single-center study including 81 nAMD patients (116 eyes), those both newly diagnosed and switched from ranibizumab. All patients had a follow-up duration of at least seven years. Treatment involved three monthly injections followed by either a pro re nata (PRN) or treat and extend regimen. Follow-up care was primarily conducted by training physicians. The primary endpoint was the change in best-corrected visual acuity (BCVA) over seven years. Secondary endpoints included central retinal thickness changes, qualitative OCT parameters, macular atrophy progression, injection frequency, and treatment adherence. Results: Among the 116 eyes, 52 (44.8%) completed the seven-year follow-up. Visual acuity improved by +2.1 letters in the overall population (+6.3 letters in treatment-naive eyes) after the loading phase but gradually declined, resulting in a loss of −12.3 letters at seven years. BCVA remained stable (a loss of fewer than 15 letters) in 57.7% of eyes. Central retinal thickness (CRT) decreased significantly during follow-up in both naive and switcher eyes. Macular atrophy occurred in 94.2% of eyes, progressing from 1.42 mm² to 8.55 mm² over seven years (p < 0.001). The mean number of injections was 4.1 ± 1.8 during the first year and 3.7 per year thereafter. Advanced age at diagnosis was a risk factor for loss to follow-up, with bilaterality being a protective factor against loss to follow-up (p < 0.05). Conclusions: This study highlights the challenges faced by a retina clinic in a teaching hospital. Suboptimal functional and anatomical outcomes in real life may derive from insufficient patient information and inconsistent monitoring, which contributes to undertreatment and affects long-term visual outcomes. It also raises concerns about supervision in a teaching hospital which needs to be improved. Full article

Visual field (VF) testing remains a cornerstone in assessing retinal function by measuring how well different parts of the retina detect light. It is essential for early detection, monitoring, and management of many retinal diseases. By mapping retinal sensitivity, VF exams can reveal functional loss before structural changes become visible. This review summarizes how VF testing is applied across key conditions: hydroxychloroquine (HCQ) retinopathy, age-related macular degeneration (AMD), diabetic retinopathy (DR) and macular edema (DME), and inherited disorders including inherited dystrophies such as retinitis pigmentosa (RP). Traditional methods like the Goldmann kinetic perimetry and simple tools such as the Amsler grid help identify large or central VF defects. Automated perimetry (e.g., Humphrey Field Analyzer) provides detailed, quantitative data critical for detecting subtle paracentral scotomas in HCQ retinopathy and central vision loss in AMD. Frequency-doubling technology (FDT) reveals early neural deficits in DR before blood vessel changes appear. Microperimetry offers precise, localized sensitivity maps for macular diseases. Despite its value, VF testing faces challenges including patient fatigue, variability in responses, and interpretation of unreliable results. Recent advances in artificial intelligence, virtual reality perimetry, and home-based perimetry systems are improving test accuracy, accessibility, and patient engagement. Integrating VF exams with these emerging technologies promises more personalized care, earlier intervention, and better long-term outcomes for patients with retinal disease. Full article

Background/Objectives: This study evaluates whether ultrawide-field optical coherence tomography angiography (UWF-OCTA) can guide navigated laser therapy for non-perfused areas (NPAs) in branch retinal vein occlusion (BRVO). It further explores whether the laser spots can be accurately placed according to plan, considering that the retina is three-dimensional (3D), while UWF-OCTA provides two-dimensional (2D) images. Methods: UWF-OCTA images from three devices—VG200, Xephilio OCT-S1, and Bmizar—guided the treatments. These images were superimposed onto NAVILAS^® system images to guide NPA treatments. Pre-treatment planning was strategically designed to avoid normal and collateral vessels, with immediate post-laser OCTA and en face images assessing the efficacy of the laser spots in avoiding these vessels as planned. The accuracy of navigated laser therapy was further analyzed by comparing the intended laser locations with the actual spots. Results: All montaged OCTA images from the three devices were seamlessly integrated into the navigated laser system without registration errors. All patients received treatments targeting the NPAs as planned. However, not all collateral or normal vessels were successfully avoided by the laser spots. A further analysis revealed that the actual locations of the laser spots deviated slightly from the planned locations, particularly in the mid-periphery areas. Conclusions: UWF-OCTA-guided navigated laser photocoagulation is feasible and precise for treating NPAs in BRVO. Nonetheless, minor deviations between planned and actual locations were observed. This discrepancy, particularly important when treating diseases of the macular area, should be carefully considered when employing OCTA-guided navigated laser photocoagulation. Full article

Adaptive optics scanning laser ophthalmoscopy optoretinography quantifies cellular function in the living retina by measuring the en face intensity change in cone photoreceptors due to visual stimulation. To fulfill the potential of optoretinography as a biomarker for assessing function in disease, we require normative optoretinographic measurements across the retina. Here we provide such measurements. We use a custom adaptive optics scanning laser ophthalmoscope to investigate cone optoretinogram (ORG) amplitudes across retinal eccentricity in five normal-sighted participants. For this purpose, we aggregated signals across cones in each measurement (~1° by 1° patch) to provide a measurement we call the population ORG. Average population ORG amplitudes decreased with increasing eccentricity for all participants, although there were individual differences in the detailed pattern of the decrease. ORG amplitudes were correlated with the thickness of the outer retina as measured using clinical OCT, which also decreases with eccentricity. Characterizing the population cone ORG as a function of eccentricity in normal-sighted participants is an important step towards establishing norms that will allow it to be used as a biomarker for assessing photoreceptor function in retinal disease. Full article

Background: Virtual clinics (VCs) have proven to be an effective solution for the increasing strain on Medical Retina (MR) services, although imaging quality issues (IQIs) persist. Our aim was to compare the quality of two ultra-wide-field (UWF) imaging modalities (Optos and Clarus) in real-world MR-VC settings. Methods: We conducted a real-world, prospective study. Data were collected from 6 Moorfields NHS Trust MR-VCs between September and October 2024. We obtained patient demographics and characteristics, primary diagnosis, UWF imaging types and images obtained, and follow-up outcomes. Results: Optos (California RG/RGB, and Monaco) was used for 56.7% (n = 152) and Zeiss Clarus 500 for 43.3% (n = 116) of the total cohort (n = 268). No statistically significant difference (p = 0.14) was found between the two in terms of the rates of IQIs. FAF (p = 0.001) acquisition was significantly higher when Optos was used. Of the patients affected by IQIs, 10 were examined in a face-to-face clinic (F2FC). No difference in IQI rates was observed when pathology-related poor image quality was considered (p = 0.561). A significantly (p = 0.001) higher rate of F2F follow-ups was found for red-flag pathologies and unexplained vision loss, with a statistically significantly higher rate of virtual follow-ups for non-red-flag pathologies (p = 0.001). Conclusions: A total of 7.5% of the clinical decisions were impacted by IQIs; 11.1% of F2FC follow-ups. Neither UWF imaging modality type was inferior in terms of IQI rates. FAF acquisition was higher with Optos, likely representing greater user-dependency for Clarus. The outcomes were not influenced by FAF acquisition, indicating that routine acquisition is not required in MR-VCs and instead should be obtained when clinically required. Full article

For surveillance video management in university laboratories, issues such as occlusion and low-resolution face capture often arise. Traditional face recognition algorithms are typically static and rely heavily on clear images, resulting in inaccurate recognition for low-resolution, small-sized faces. To address the challenges of occlusion and low-resolution person identification, this paper proposes a new face recognition framework by reconstructing Retinaface-Resnet and combining it with Quality-Adaptive Margin (adaface). Currently, although there are many target detection algorithms, they all require a large amount of data for training. However, datasets for low-resolution face detection are scarce, leading to poor detection performance of the models. This paper aims to solve Retinaface’s weak face recognition capability in low-resolution scenarios and its potential inaccuracies in face bounding box localization when faces are at extreme angles or partially occluded. To this end, Spatial Depth-wise Separable Convolutions are introduced. Retinaface-Resnet is designed for face detection and localization, while adaface is employed to address low-resolution face recognition by using feature norm approximation to estimate image quality and applying an adaptive margin function. Additionally, a multi-object tracking algorithm is used to solve the problem of moving occlusion. Experimental results demonstrate significant improvements, achieving an accuracy of 96.12% on the WiderFace dataset and a recognition accuracy of 84.36% in practical laboratory applications. Full article