Search Results (798)

Retinal vessel segmentation in fundus images is critical for diagnosing microvascular and ophthalmologic diseases. However, the task remains challenging due to significant vessel width variation and low vessel-to-background contrast. To address these limitations, we propose WDM-UNet, a novel spatial-wavelet dual-domain fusion architecture that integrates spatial and wavelet-domain representations to simultaneously enhance the local detail and global context. The encoder combines a Deformable Convolution Encoder (DCE), which adaptively models complex vascular structures through dynamic receptive fields, and a Wavelet Convolution Encoder (WCE), which captures the semantic and structural contexts through low-frequency components and hierarchical wavelet convolution. These features are further refined by a Gated Fusion Transformer (GFT), which employs gated attention to enhance multi-scale feature integration. In the decoder, depthwise separable convolutions are used to reduce the computational overhead without compromising the representational capacity. To preserve fine structural details and facilitate contextual information flow across layers, the model incorporates skip connections with a hierarchical fusion strategy, enabling the effective integration of shallow and deep features. We evaluated WDM-UNet in three public datasets: DRIVE, STARE, and CHASE_DB1. The quantitative evaluations demonstrate that WDM-UNet consistently outperforms state-of-the-art methods, achieving 96.92% accuracy, 83.61% sensitivity, and an 82.87% F1-score in the DRIVE dataset, with superior performance across all the benchmark datasets in both segmentation accuracy and robustness, particularly in complex vascular scenarios. Full article

Multiple sclerosis (MS), a chronic disease of the central nervous system, is known to cause structural and vascular changes in the retina. Although optical coherence tomography (OCT) and fundus photography can detect retinal thinning and circulatory abnormalities, these findings are not specific to MS. This study explores the potential of Infrared Scanning-Laser-Ophthalmoscopy (IR-SLO) imaging to uncover vascular morphological features that may serve as MS-specific biomarkers. Using an age-matched, subject-wise stratified k-fold cross-validation approach, a deep learning model originally designed for color fundus images was adapted to segment optic disc, optic cup, and retinal vessels in IR-SLO images, achieving Dice coefficients of 91%, 94.5%, and 97%, respectively. This process included tailored pre- and post-processing steps to optimize segmentation accuracy. Subsequently, clinically relevant features were extracted. Statistical analyses followed by SHapley Additive exPlanations (SHAP) identified vessel fractal dimension, vessel density in zones B and C (circular regions extending 0.5–1 and 0.5–2 optic disc diameters from the optic disc margin, respectively), along with vessel intensity and width, as key differentiators between MS patients and healthy controls. These findings suggest that IR-SLO can non-invasively detect retinal vascular biomarkers that may serve as additional or alternative diagnostic markers for MS diagnosis, complementing current invasive procedures. Full article

The eye, a complex organ essential for visual perception, is composed of diverse cell populations with specialized functions; however, the complex interplay between these cellular components and their underlying molecular mechanisms remains largely elusive. Traditional biotechnologies, such as bulk RNA sequencing and in vitro models, are limited in capturing cellular heterogeneity or accurately mimicking the complexity of human ophthalmic diseases. The advent of single-cell RNA sequencing (scRNA-seq) has revolutionized ocular research by enabling high-resolution analysis at the single-cell level, uncovering cellular heterogeneity, and identifying disease-specific gene profiles. In this review, we provide a review of scRNA-seq application advancement in ocular physiology and pathology, highlighting its role in elucidating the molecular mechanisms of various ocular diseases, including myopia, ocular surface and corneal diseases, glaucoma, uveitis, retinal diseases, and ocular tumors. By providing novel insights into cellular diversity, gene expression dynamics, and cell–cell interactions, scRNA-seq has facilitated the identification of novel biomarkers and therapeutic targets, and the further integration of scRNA-seq with other omics technologies holds promise for deepening our understanding of ocular health and diseases. Full article

Age-related macular degeneration (AMD) is the leading cause of blindness in developed countries. Predicting its progression is crucial for preventing late-stage AMD, as it is an irreversible retinal disease. Both genetic factors and retinal images are instrumental in diagnosing and predicting AMD progression. Previous studies have explored automated diagnosis using single fundus images and genetic variants, but they often fail to utilize the valuable longitudinal data from multiple visits. Longitudinal retinal images offer a dynamic view of disease progression, yet standard Long Short-Term Memory (LSTM) models assume consistent time intervals between training and testing, limiting their effectiveness in real-world settings. To address this limitation, we propose time-varied Long Short-Term Memory (TV-LSTM), which accommodates irregular time intervals in longitudinal data. Our innovative approach enables the integration of both longitudinal fundus images and AMD-associated genetic variants for more precise progression prediction. Our TV-LSTM model achieved an AUC-ROC of 0.9479 and an AUC-PR of 0.8591 for predicting late AMD within two years, using data from four visits with varying time intervals. Full article

The retina offers a unique window into both ocular and systemic health, motivating the development of AI-based tools for disease screening and risk assessment. In this study, we present a comprehensive evaluation of six state-of-the-art deep neural networks, including convolutional neural networks and vision transformer architectures, on the Brazilian Multilabel Ophthalmological Dataset (BRSET), comprising 16,266 fundus images annotated for multiple clinical and demographic labels. We explored seven classification tasks: Diabetes, Diabetic Retinopathy (2-class), Diabetic Retinopathy (3-class), Hypertension, Hypertensive Retinopathy, Drusen, and Sex classification. Models were evaluated using precision, recall, F1-score, accuracy, and AUC. Among all models, the Swin-L generally delivered the best performance across scenarios for Diabetes (AUC = 0.88, weighted F1-score = 0.86), Diabetic Retinopathy (2-class) (AUC = 0.98, weighted F1-score = 0.95), Diabetic Retinopathy (3-class) (macro AUC = 0.98, weighted F1-score = 0.95), Hypertension (AUC = 0.85, weighted F1-score = 0.79), Hypertensive Retinopathy (AUC = 0.81, weighted F1-score = 0.97), Drusen detection (AUC = 0.93, weighted F1-score = 0.90), and Sex classification (AUC = 0.87, weighted F1-score = 0.80). These results reflect excellent to outstanding diagnostic performance. We also employed gradient-based saliency maps to enhance explainability and visualize decision-relevant retinal features. Our findings underscore the potential of deep learning, particularly vision transformer models, to deliver accurate, interpretable, and clinically meaningful screening tools for retinal and systemic disease detection. Full article

Background and aim: Advanced Therapy Medicinal Products (ATMPs) are innovative drugs based on genes, tissues, or cells that target rare and severe diseases. ATMPs have shown promising clinical outcomes but are associated with high costs, raising questions about cost-effectiveness. Hence, this systematic review aims to analyze the cost-effectiveness and cost-utility profiles of the European Medicines Agency-authorized ATMPs for treating rare diseases. Methods: A systematic review was conducted following PRISMA guidelines. Studies were identified by searching PubMed, Embase, Web of Science, and ProQuest scientific databases. Economic evaluations reporting incremental cost-effectiveness/utility ratios (ICERs/ICURs) for ATMPs were included. Costs were standardized to 2023 Euros, and a cost-effectiveness plane was constructed to evaluate the results against willingness-to-pay (WTP) thresholds of EUR 50,000, EUR 100,000, and EUR 150,000 per QALY, as part of a sensitivity analysis. Results: A total of 61 studies met the inclusion criteria. ATMPs for rare blood diseases, such as tisagenlecleucel and axicabtagene ciloleucel, were found to be cost-effective in a majority of studies, with incremental QALYs ranging from 1.5 to 10 per patient over lifetime horizon. Tisagenlecleucel demonstrated a positive cost-effectiveness profile in the treatment of acute lymphoblastic leukemia (58%), while axicabtagene ciloleucel showed a positive profile in the treatment of diffuse large B-cell lymphoma (85%). Onasemnogene abeparvovec for spinal muscular atrophy (SMA) showed uncertain cost-effectiveness results, and voretigene neparvovec for retinal diseases was not cost-effective in 40% of studies, with incremental QALYs around 1.3 and high costs exceeding the WTP threshold set. Conclusions: ATMPs in treating rare diseases show promising economic potential, but cost-effectiveness varies across indications. Policymakers must balance innovation with system sustainability, using refined models and the long-term impact on patient outcomes. Full article

Marine ecosystems are characterized by an immense biodiversity and represent a rich source of biological compounds with promising potential for the development of novel therapeutic drugs. This review describes the most promising marine-derived neuroprotective compounds with strong potential for the treatment of neurodegenerative disorders. We focus specifically on the retina and brain—two key components of the central nervous system—as primary targets for therapeutic interventions against neurodegeneration. Alzheimer’s disease and retinal degeneration diseases are used here as a representative model of neurodegenerative disorders, where complex molecular processes such as protein misfolding, oxidative stress, and neuroinflammation drive disease progression. We also examine gene therapy approaches inspired by marine biology, with particular attention to their application in retinal diseases, aimed at preserving or restoring photoreceptor function and vision. Full article

Background: Retinal diseases are becoming an important public health issue, with early diagnosis and timely intervention playing a key role in preventing vision loss. Optical coherence tomography (OCT) remains the leading non-invasive imaging technique for identifying retinal conditions. However, distinguishing between diabetic macular edema (DME) and macular edema resulting from retinal vein occlusion (RVO) can be particularly challenging, especially for clinicians without specialized training in retinal disorders, as both conditions manifest through increased retinal thickness. Due to the limited research exploring the application of deep learning methods, particularly for RVO detection using OCT scans, this study proposes a novel diagnostic approach based on stacked convolutional neural networks. This architecture aims to enhance classification accuracy by integrating multiple neural network layers, enabling more robust feature extraction and improved differentiation between retinal pathologies. Methods: The VGG-16, VGG-19, and ResNet50 models were fine-tuned using the Kermany dataset to classify the OCT images and afterwards were trained using a private OCT dataset. Four stacked models were then developed using these models: a model using the VGG-16 and VGG-19 networks, a model using the VGG-16 and ResNet50 networks, a model using the VGG-19 and ResNet50 models, and finally a model using all three networks. The performance metrics of the model includes accuracy, precision, recall, $F_2$-score, and area under of the receiver operating characteristic curve (AUROC). Results: The stacked neural network using all three models achieved the best results, having an accuracy of 90.7%, precision of 99.2%, a recall of 90.7%, and an $F_2$-score of 92.3%. Conclusions: This study presents a novel method for distinguishing retinal disease by using stacked neural networks. This research aims to provide a reliable tool for ophthalmologists to improve diagnosis accuracy and speed. Full article

The retinal pigment epithelium (RPE), a monolayer of pigmented cells, is critical for visual function through its interaction with the neural retina. In healthy eyes, RPE cells exhibit a uniform hexagonal arrangement, but under stress or disease, such as age-related macular degeneration (AMD), dysmorphic traits like cell enlargement and apparent multinucleation emerge. Multinucleation has been hypothesized to result from cellular fusion, a compensatory mechanism to maintain cell-to-cell contact and barrier function, as well as conserve resources in unhealthy tissue. However, traditional two-dimensional (2D) imaging using apical border markers alone may misrepresent multinucleation due to the lack of lateral markers. We present high-resolution confocal images enabling three-dimensional (3D) visualization of apical (ZO-1) and lateral (α-catenin) markers alongside nuclei. In two RPE damage models, we find that seemingly multinucleated cells are often single cells with displaced neighboring nuclei and lateral membranes. This emphasizes the need for 3D analyses to avoid misidentifying multinucleation and underlying fusion mechanisms. Lastly, images from the NaIO₃ oxidative damage model reveal variability in RPE damage, with elongated, dysmorphic cells showing increased ZsGreen reporter protein expression driven by EMT-linked CAG promoter activity, while more regular RPE cells displayed somewhat reduced green signal more typical of epithelial phenotypes. Full article

Age-related ocular conditions like macular degeneration (AMD), diabetic retinopathy (DR), and glaucoma are leading causes of irreversible vision loss globally. Optical coherence tomography (OCT) provides essential non-invasive visualization of retinal structures for early diagnosis, but manual analysis of these images is labor-intensive and prone to variability. Deep learning (DL) techniques have emerged as powerful tools for automating the segmentation of the retinal layer in OCT scans, potentially improving diagnostic efficiency and consistency. This review systematically evaluates the state of the art in DL-based retinal layer segmentation using the PRISMA methodology. We analyze various architectures (including CNNs, U-Net variants, GANs, and transformers), examine the characteristics and availability of datasets, discuss common preprocessing and data augmentation strategies, identify frequently targeted retinal layers, and compare performance evaluation metrics across studies. Our synthesis highlights significant progress, particularly with U-Net-based models, which often achieve Dice scores exceeding 0.90 for well-defined layers, such as the retinal pigment epithelium (RPE). However, it also identifies ongoing challenges, including dataset heterogeneity, inconsistent evaluation protocols, difficulties in segmenting specific layers (e.g., OPL, RNFL), and the need for improved clinical integration. This review provides a comprehensive overview of current strengths, limitations, and future directions to guide research towards more robust and clinically applicable automated segmentation tools for enhanced ocular disease diagnosis. Full article

Background: Optical Coherence Tomography has become a common imaging technique that enables a non-invasive and detailed visualization of the retina and allows for the identification of various diseases. Through the advancement of technology, the volume and complexity of OCT data have rendered manual analysis infeasible, creating the need for automated means of detection. Methods: This study investigates the ability of state-of-the-art object detection models, including the latest YOLO versions (from v8 to v12), YOLO-World, YOLOE, and RT-DETR, to accurately detect pathological biomarkers in two retinal OCT datasets. The AROI dataset focuses on fluid detection in Age-related Macular Degeneration, while the OCT5k dataset contains a wide range of retinal pathologies. Results: The experiments performed show that YOLOv12 offers the best balance between detection accuracy and computational efficiency, while YOLOE manages to consistently outperform all other models across both datasets and most classes, particularly in detecting pathologies that cover a smaller area. Conclusions: This work provides a comprehensive benchmark of the capabilities of state-of-the-art object detection for medical applications, specifically for identifying retinal pathologies from OCT scans, offering insights and a starting point for the development of future automated solutions for analysis in a clinical setting. Full article

Ocular diseases can significantly affect vision and overall quality of life, with diagnosis often being time-consuming and dependent on expert interpretation. While previous computer-aided diagnostic systems have focused primarily on medical imaging, this paper proposes VisionTrack, a multi-modal AI system for predicting multiple retinal diseases, including Diabetic Retinopathy (DR), Age-related Macular Degeneration (AMD), Diabetic Macular Edema (DME), drusen, Central Serous Retinopathy (CSR), and Macular Hole (MH), as well as normal cases. The proposed framework integrates a Convolutional Neural Network (CNN) for image-based feature extraction, a Graph Neural Network (GNN) to model complex relationships among clinical risk factors, and a Large Language Model (LLM) to process patient medical reports. By leveraging diverse data sources, VisionTrack improves prediction accuracy and offers a more comprehensive assessment of retinal health. Experimental results demonstrate the effectiveness of this hybrid system, highlighting its potential for early detection, risk assessment, and personalized ophthalmic care. Experiments were conducted using two publicly available datasets, RetinalOCT and RFMID, which provide diverse retinal imaging modalities: OCT images and fundus images, respectively. The proposed multi-modal AI system demonstrated strong performance in multi-label disease prediction. On the RetinalOCT dataset, the model achieved an accuracy of 0.980, F1-score of 0.979, recall of 0.978, and precision of 0.979. Similarly, on the RFMID dataset, it reached an accuracy of 0.989, F1-score of 0.881, recall of 0.866, and precision of 0.897. These results confirm the robustness, reliability, and generalization capability of the proposed approach across different imaging modalities. Full article

Retinal ischemia is implicated in ocular diseases involving aberrant neovessel proliferation that characterizes proliferative retinopathies. Their therapy still remains confined to the intravitreal administration of anti-vascular endothelial growth factor (VEGF) medication, which is limited by side effects and progressive reduction in efficacy. Mimicking neovascular diseases in rodents, although of great help for translating fundamental mechanistic findings and assessing therapeutic potential in humans, is limited by the rodent’s short life span, which prevents retinal vessel proliferation over time. However, the oxygen-induced retinopathy (OIR) model, which mimics retinopathy of prematurity, seems to meet some criteria that are common to proliferative retinopathies. The present review provides insight into preclinical models and their suitability to mimic proliferative retinopathies. Further considerations will be applied to emerging approaches and advanced methodologies for the management of proliferative retinopathies, leading to the identification of new therapeutic targets, including our contribution in the field. Major emphasis is given to the possibility of using systemic therapies either alone or in combination with intravitreal anti-VEGF administration to maximize clinical benefits by combining drugs with different modes of action. This review is concluded by an in-depth discussion on future advancements and a critical view of preclinical finding translatability. Despite the major effort of preclinical and clinical research to develop novel therapies, the blockade of VEGF activity still remains the only treatment for proliferative retinopathies for more than twenty years since its first therapeutic application. Full article

Bestrophinopathies are a group of inherited retinal diseases caused by mutations in the BEST1 gene. The protein encoded by this gene, bestorphin-1 (hBest1), is a calcium-dependent transmembrane channel localized on the basolateral membrane of retinal pigment epithelial (RPE) cells. We have already demonstrated the surface behavior and organization of recombinant hBest1 and its interactions with membrane lipids such as 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), sphingomyelin (SM) and cholesterol (Chol) in models of biological membranes, which affect the hBest1 structure–function relationship. The main aim of our current investigation is to integrate pure hBest1 protein into lipid bilayer nanostructures. We synthesized and characterized various hBest1-containing nanostructures based on 1,2-Dipalmitoylphosphatidylcholine (DPPC), SM, glycerol monooleate (GMO) and Chol in different ratios and determined their cytotoxicity and incorporation into cell membranes and/or cells by immunofluorescence staining. Our results show that these newly designed nanoparticles are not cytotoxic and that their incorporation into MDCK II cell membranes (used as a model system) may provide a mechanism that could be applied to RPE cells expressing mutated hBest1 in order to restore their ion transport functions, affected by mutated and malfunctioning hBest1 molecules. Full article

Background/Objective: The Rs1 exon-1-del rat (Rs1KO) XLRS model shows normal retinal development until postnatal day 12 (P12) when small cystic spaces start to form in the inner nuclear layer. These enlarge rapidly, peak at P15, and then collapse by P19. These events overlap with eye opening at P12–P15. We investigated whether new light-driven retinal activity could contribute to the appearance and progression of schisis cavities in this rat model of XLRS disease. Methods: For dark rearing (D/D), mating pairs of Rs1KO strain were raised in total darkness in a special vivarium at UC Davis. When pups were born, they were maintained in total darkness, and eyes were collected at P12, P15, and P30 (n = 3/group) for each of the D/D and cyclic light-reared 12 h light–12 h dark (L/D) Rs1KO and wild-type (WT) littermates. Eyes were fixed, paraffin-embedded, and sectioned. Tissue morphology was examined by H&E and marker expression of retinoschisin1 (Rs1), rhodopsin (Rho), and postsynaptic protein 95 (Psd95) by fluorescent immunohistochemistry. H&E-stained images were analyzed with ImageJ version 1.54h to quantify cavity size using the “Analyze Particles” function. Results: Small intra-retinal schisis cavities begin to form by P12 in the inner retina of both D/D and L/D animals. Cavity formation was equivalent or more pronounced in D/D animals than in L/D animals. We compared Iba1 (activation marker of immune cells) distribution and found that by P12, when schisis appeared, Iba1⁺ cells had accumulated in regions of schisis. Iba1⁺ cells were more abundant in Rs1KO animals than WT animals and appeared slightly more prevalent in D/D- than L/D-reared Rs1KO animals. We compared photoreceptor development using Rho, Rs1, and Psd95 expression, and these were similar; however, the outer segments (OSs) of D/D animals with Rho labeling at P12 were longer than L/D animals. Conclusions: The results showed that cavities formed at the same time in D/D and L/D XLRS rat pups, indicating that the timing of schisis formation is not light stimulus-driven but rather appears to be a result of developmental events. Cavity size tended to be larger under dark-rearing conditions in D/D animals, which could be due to the decreased rate of phagocytosis by the RPE in the dark, allowing for continued growth of the OSs without the usual shedding of the distal tip, a key mechanism behind dark adaptation in the retina. These results highlight the complexity of XLRS pathology; however, we found no evidence that light-driven metabolic activity accounted for schisis cavity formation. Full article