Search Results (429)

To perform screening of the retina on a population scale, an automated procedure is required that incorporates accurate, reproducible, interpretable, and computationally costeffective models. Existing approaches using convolutional or transformer architectures typically do not adequately represent both fine-grained pathology and large-scale retinal context simultaneously, which could adversely affect their reliability if used for large-scale applications in clinical practice. In this paper, we propose a hierarchical transformer-based screening framework for retinal fundus images that incorporates patch-based tokenization, global transformer encoding, and hierarchical aggregation of contextual information. We also developed a lightweight prediction head that supports screening for both single and multiple diseases. The framework has been evaluated using standard screening metrics, robustness, and cross-dataset generalization analyses on two eye retinopathy image databases: EyePACS and RFMiD. With regard to screening for a binary outcome of diabetic retinopathy, our method provided an accuracy of 89.4% and an area under the receiver operating characteristic (AUROC) curve of 93.6% on EyePACS and attained an accuracy of 95.2% and a macro-averaged F1 score of 82.7% on RFMiD. Our hierarchical transformer achieved improved robustness to degraded images and increased generalizability across datasets compared with all current state-of-the-art models. The proposed hierarchical transformer demonstrates strong potential for large-scale retinal screening and provides a promising foundation for future clinically validated deployment. Full article

Background/Objectives: Optical coherence tomography angiography (OCTA) enables noninvasive quantitative assessment of the retinal microvasculature and is widely used in diabetic retinopathy (DR). However, OCTA-derived metrics are highly dependent on post-processing techniques, particularly vessel binarization. This study aimed to compare local and global binarization methods applied after Frangi filtering for vessel enhancement in parafoveal vessel density analysis. Methods: This cross-sectional study included 69 participants: 17 healthy controls and 52 diabetic patients, classified as the following: no DR (n = 14), non-proliferative DR (NPDR, n = 18), or proliferative DR (PDR, n = 20). All subjects underwent comprehensive ophthalmological examination and OCTA imaging of the superficial capillary plexus using a Topcon OCTA system. Images were processed using a custom MATLAB protocol. Following Frangi filtering, five binarization methods were applied: three local (Phansalkar, local Otsu, adaptive mean) and two global (global mean and global Otsu). Parafoveal vessel density was quantified within the four inner quadrants of the ETDRS grid. Results: Statistically significant differences in vessel density were consistently observed between PDR group and both the control and no DR groups across all local binarization methods. Among global methods, only global Otsu thresholding detected a significant difference between PDR and control. The most robust differences were predominantly identified in the nasal and inferior quadrants. Conclusions: Local adaptive binarization methods demonstrated superior sensitivity and structural preservation for parafoveal vessel density analysis in DR. Global methods showed limited discriminative capability. These findings support the preferential use of local adaptive techniques for reliable OCTA-based vascular assessment in diabetic retinopathy. Full article

Access to the sky is a key element of residential environmental quality. In densely built-up urban areas, exposure to the sky is often limited not only quantitatively but, above all, directionally. Traditional illuminance metrics, such as the Sky View Factor (SVF) or Daylight Factor (DF), describe the proportion of visible sky or the amount of light in an averaged manner, without considering its relationship to the functional organisation of the human field of view.This article introduces the Relative Retinal Image (RRI) metric, which evaluates directional access to the sky through geometric analysis of viewing directions in relation to functional zones of the visual field, without reconstructing perceived images or simulating physiological processes. Within this geometric framework, human vision is interpreted as operating simultaneously in two visual cones: a narrow central cone responsible for acute, conscious vision (RRI-A), and a wider peripheral cone enabling the reception of low-resolution but spatially stable stimuli (RRI-B). For clarity, three concentric central ranges are distinguished: foveal (0–2.5°), sharp central (0–5°), and extended interpretative central vision (up to 10°). The proposed approach provides a geometry-based analytical tool that complements existing daylight metrics in the assessment of sustainable residential environments, without formulating normative or biological design prescriptions. Based on geometric and graphical analyses and a case study of the Józefowiec housing estate in Katowice, the results indicate that the directional structure of the sky view may be lost despite compliance with conventional planning criteria. Full article

Background/Objectives: We aimed to evaluate the diagnostic accuracy of the MONA.health artificial intelligence (AI) software (Version 1.0.0; MONA.health, Leuven, Belgium) and compare its advantages in screening for diabetic retinopathy (DR) and diabetic macular edema (DME) with standard fundus photography. Methods: This cross-sectional, real-life instrument validation study was conducted at the Vuk Vrhovac University Clinic in Zagreb during routine DR screening and included 296 patients (592 eyes) with diabetes. Following standard fundus photography using a 45° Zeiss VISUCAM NM/FA camera (Carl Zeiss Meditec AG, Jena, Germany), each patient also underwent imaging with an automated portable retinal camera (NFC-600, Crystalvue Ophthalmic Instruments, Taoyuan City, Taiwan). Two retina specialists independently graded images from the standard camera, while images from the NFC-600 were analyzed using the MONA.health AI software. Results: Among the 592 eyes, human grading identified 81 with any DR, including 17 with mild NPDR, 64 with referable DR (moderate/severe NPDR or PDR), and 13 with DME. The MONA.health AI software identified 65 eyes with referable DR and 19 with DME. For MONA DR screening compared to the standard fundus camera, the area under the curve, sensitivity, specificity, positive predictive value, negative predictive value, positive likelihood ratio, negative likelihood ratio, kappa agreement, diagnostic odds ratio, and diagnostic effectiveness were 99.74%, 100%, 99.81%, 99.33%, 100%, 528.00, 0.00, 0.99, infinity, and 99.85%, respectively. For MONA DME screening, these metrics were 97.97%, 100%, 98.95%, 85.93%, 100%, 95.67, 0.00, 0.81, infinity, and 99.02%, respectively. The MONA AI screening process required 1 day of training and approximately 5 min for image capture and analysis, compared to 7 days of training and 13 min for image acquisition and grading with the standard method. Conclusions: These findings demonstrate that the MONA.health AI software matches the accuracy of standard fundus photography for screening and early detection of referable DR and DME, while offering a faster, simpler, and more user-friendly workflow that significantly reduces the time to obtain screening results. Full article

Medical image classification is a fundamental task in medical image analysis and underpins a wide range of clinical applications, including dermatological screening, retinal disease assessment, and malignant tissue detection. In recent years, diffusion models have demonstrated promising potential for medical image classification owing to their strong representation learning capability. However, existing diffusion-based classification methods often rely on oversimplified prior modeling strategies, which fail to adequately capture the intrinsic multi-scale semantic information and contextual dependencies inherent in medical images. As a result, the discriminative power and stability of feature representations are constrained in complex scenarios. In addition, fixed noise injection strategies neglect variations in sample-level prediction confidence, leading to uniform perturbations being imposed on samples with different levels of semantic reliability during the diffusion process, which in turn limits the model’s discriminative performance and generalization ability. To address these challenges, this paper proposes a Confidence-Guided Adaptive Diffusion Network (CGAD-Net) for medical image classification. Specifically, a hybrid prior modeling framework is introduced, consisting of a Hierarchical Pyramid Context Modeling (HPCM) module and an Intra-Scale Dilated Convolution Refinement (IDCR) module. These two components jointly enable the diffusion-based feature modeling process to effectively capture fine-grained structural details and global contextual semantic information. Furthermore, a Confidence-Guided Adaptive Noise Injection (CG-ANI) strategy is designed to dynamically regulate noise intensity during the diffusion process according to sample-level prediction confidence. Without altering the underlying discriminative objective, CG-ANI stabilizes model training and enhances robust representation learning for semantically ambiguous samples.Experimental results on multiple public medical image classification benchmarks, including HAM10000, APTOS2019, and Chaoyang, demonstrate that CGAD-Net achieves competitive performance in terms of classification accuracy, robustness, and training stability. These results validate the effectiveness and application potential of confidence-guided diffusion modeling for two-dimensional medical image classification tasks, and provide valuable insights for further research on diffusion models in the field of medical image analysis. Full article

Background/Objectives: Coronary artery disease (CAD) remains one of the leading cardiovascular diseases worldwide. While obstructive CAD is well characterized and managed, identification of patients with non-obstructive CAD (NOCAD) remains challenging. Unlike the coronary vasculature, the eye’s microcirculation can be easily and non-invasively assessed. Therefore, this systematic review summarized the ophthalmological diagnostic methods used to assess microvascular alterations associated with coronary microvascular dysfunction (CMD), angina with non-obstructive coronary arteries (ANOCA), or ischemia with non-obstructive coronary arteries (INOCA). Methods: According to PRISMA guidelines, PubMed/MEDLINE and Embase databases were screened by two independent reviewers from inception to 25 November 2025. Original articles that examined ophthalmological microvascular changes by any method in adults with CMD or its subtypes were included. The quality of the studies was assessed using the JBI Critical Appraisal Checklist. Results: Of 101 identified articles, nine studies met the inclusion criteria, comprising 1894 patients. Optical coherence tomography angiography was the most frequently used imaging modality, followed by optical coherence tomography, slit-lamp smartphone imaging, and fundus photography. Five investigations employed blinded image analysis, three did not, and one study used it partially. Four studies used semi-automated measurements, four employed fully automated methods, and one study applied manual and automated measurements for different parameters. Conclusions: Despite a limited number of studies, retinal and conjunctival microvascular alterations helped differentiate CAD subtypes and may reflect systemic microcirculatory impairment among patients with ANOCA/INOCA. Ophthalmological imaging techniques have the potential to serve as non-invasive tools for detecting microvascular alterations associated with CMD in ANOCA and INOCA patients. PROSPERO Registration Number: CRD420251239875 Full article

Background/Objectives: To investigate the rate of exudative progression over time in patients with non-exudative macular neovascularization (NE-MNV) associated with various acquired macular degenerations presenting with a double-layer sign (DLS) or triple-layer sign (TLS) on optical coherence tomography (OCT), and to identify potential predictors of this progression. Methods: Fifty-one eyes of fourty-nine patients with a DLS or TLS on OCT images were identified. OCT angiography (OCTA) was performed to detect NE-MNV, and only eyes with confirmed NE-MNV were included in the final analysis. Central macular thickness (CMT), choroidal thickness (CT), morphology of the abnormal vessels, the duration of follow-up, progression to active exudative MNV, and the status of the contralateral eye were assessed. Results: The final analysis included 32 eyes of 30 participants with NE-MNV. The median observation period was 46 months. The causes of NE-MNV were age- related macular degeneration (AMD) in 59.38% of eyes, pachychoroid epitheliopathy (PPE) in 37.50%, and other causes in 3.12%. Exudation developed in 15.62% of eyes (median time to onset: 24 months), predominantly in the AMD subgroup. Abnormalities in the fellow eye were present in 59.38% of cases. Neither age nor other factors, including sex, cause of MNV, CMT, CT, MNV morphology, or fellow eye status, were statistically significant predictors of progression to active MNV (p = 0.67, p > 0.99, p = 0.62, p = 0.09, p = 0.09, p = 0.2, p = 0.62, resp.). Conclusions: NE-MNV is an asymptomatic condition that may occur in the course of various retinal diseases. While DLS and TLS demonstrate high sensitivity and specificity for the diagnosis of NE-MNV, their presence does not always indicate concurrent MNV. Multimodal imaging is essential for accurate monitoring of these patients and detection of potential disease progression. Full article

This study examines the efficacy of phasor analysis in distinguishing between healthy and diseased retinas using multispectral imaging data together with machine learning approaches. Our results demonstrate that phasor analysis of multispectral images surpasses average reflectance values in classification performance, serving as an effective dimensionality reduction technique to extract essential features, with the first harmonic yielding optimal results when paired with Z-score normalization. To compare the effectiveness of multispectral images with that of a conventional color fundus camera, we extracted three spectral bands corresponding to the red, green, and blue regions and combined them to create RGB-like images, which were then subjected to the same analysis. Our study found that phasor analysis of multispectral images provided more accurate classification results than phasor analysis of RGB-like images. An examination of different regions of interest showed that using the entire retina yields the best classification performance, likely due to the advanced stage of the diseases, which had progressed to affect the entire fundus. Our findings suggest that phasor analysis of multispectral images and machine learning are a powerful tools for retinal disease classification. Full article

Background/Objectives: Diabetic retinopathy (DR) is a chronic, progressive complication of diabetes mellitus and remains one of the leading causes of vision impairment worldwide, particularly when early pathological changes go undetected or untreated. The earliest clinically identifiable biomarkers are microaneurysms, which are minute, round dilatations of capillary walls. Retinal abnormalities of a broad spectrum are indicative of the condition. This paper introduces a novel automated screening system for DR that prioritises the detection of these early indicators. Methods: The proposed approach integrates advanced image processing techniques based on the circular Hough transform and the YOLOv9 model, to localise and detect microaneurysms in colour fundus images. Results: Several system prototype versions were developed and evaluated. The final, best-performing YOLOv9-based model achieved an accuracy of 91%, representing a substantial performance improvement compared with the circular Hough transform. Conclusions: The developed models effectively address the issue of significant image processing challenges in lesion detection as well as small and class imbalance data, which are recurring constraints in medical image analysis. Full article

Optic neuritis (ON) has been recognized since antiquity, but its modern clinical identity emerged only in the late 19th century and was definitively shaped by the Optic Neuritis Treatment Trial (ONTT). The ONTT established the natural history, visual prognosis, association with multiple sclerosis (MS), and therapeutic response to corticosteroids, building the foundation for contemporary ON management. Subsequent discoveries—most notably aquaporin-4 IgG-associated ON (AQP4-ON), myelin oligodendrocyte glycoprotein antibody-associated ON (MOG-ON), and double-negative ON—have fundamentally transformed this paradigm, shifting ON from a seemingly uniform demyelinating syndrome to a group of biologically distinct disorders. These subtypes differ in immunopathology, clinical course, MRI features, retinal injury patterns, CSF profiles, and long-term outcomes, making early and accurate differentiation essential. MRI provides key distinctions in lesion length, orbital tissue inflammation, bilateral involvement, and chiasmal or optic tract extension. Optical coherence tomography (OCT) offers complementary structural biomarkers, including severe early ganglion cell loss in AQP4-ON, relative preservation in MOG-ON, and variable patterns in double-negative ON. CSF analysis further refines diagnosis, with oligoclonal bands strongly supporting MS-ON. Together, these modalities enable precise early stratification and timely initiation of targeted immunotherapy, which is critical for preventing irreversible visual disability. Despite major advances, significant unmet needs persist. Access to high-resolution MRI, OCT, cell-based antibody assays, and evidence-based treatments remains limited in many regions, contributing to global disparities in outcomes. The understanding of the pathogenesis of double-negative optic neuritis, the identification of reliable biomarkers of relapse and visual recovery, and the determination of standardized cut-off values for multimodal diagnostic tools—including MRI, OCT, CSF analysis, and serological assays—remain unresolved challenges. Future research must expand biomarker discovery, refine imaging criteria, and ensure equitable global access to cutting-edge diagnostic platforms and therapeutic innovations. Four decades after the ONTT, ON remains a dynamic field of investigation, with ongoing advances holding the potential to transform care for patients worldwide. Together, these advances expose a fundamental tension between historically MS-centered diagnostic frameworks and the emerging biological heterogeneity of ON, a tension that underpins the structure and critical perspective of the present review. Full article

Diabetic retinopathy (DR) remains the leading cause of preventable blindness among working-age adults worldwide, driven by the growing prevalence of diabetes mellitus. The aim of this comprehensive literature review is to provide an insightful analysis of recent advances in the pathogenesis of DR, followed by a summary of emerging technologies for its diagnosis and treatment. Recent studies have explored the roles of cell death pathways, immune activation, and lipid peroxidation in the pathology of DR. However, at the core of DR pathology lies neovascularization driven by vascular endothelial growth factor (VEGF), and mitochondrial damage due to dysregulated oxidative stress. These dysregulated pathways manifest clinically as DR, with specific subtypes including non-proliferative DR, proliferative DR and diabetic macular edema, which can be diagnosed through various imaging modalities. Recently, novel advances have been made using liquid biopsy and artificial (AI)-based algorithms with the goal of transforming DR diagnostics. AI models show distinct promise with the capacity to provide automated interpretation of retinal imaging. Furthermore, conventional anti-VEGF injectable agents have revolutionized DR treatment in the past decades. Today, as the pathogenesis of DR becomes better understood, new pathways, such as the ROS-VEGF loop, are being elucidated in greater depth, enabling the development of targeted therapies. In addition, new innovations such as intravitreal implants are transforming the delivery of DR-specific medication. This paper will discuss the current understanding of the pathogenesis of DR, which is leading to new diagnostic and therapeutic tools that will transform clinical management of DR. Full article

Background: The aim of this retrospective study was to evaluate correlations between Optical Coherence Tomography (OCT) and Adaptive Optics (AO) of selected retinal parameters in individuals with increased BMI (≥25.0), including a subgroup analysis for hypertension (HTN). Methods: Sixty-three patients (120 eyes) were assessed using AngioVue OCT and rtx1TM AO devices. Retinal thickness (RT), optic nerve head (ONH), ganglion cell complex (GCC), retinal nerve fiber layer (RNFL), and photoreceptor (cone) parameters—density, spacing, regularity, dispersion—were analyzed. Results: A negative correlation between BMI and RT in the parafoveal superior and inferior quadrants was observed. Higher BMI was associated with thinner GCC in the superior and nasal parafoveal regions. Additionally, age negatively correlated with cone density and regularity, and positively with cone spacing and dispersion. Numerous correlations were noted between GCC values in OCT and cone parameters in AO, consistent across both HTN and non-HTN subgroups. Conclusions: The findings suggested that AO may detect retinal changes earlier than OCT. Multimodal imaging provides valuable insights into early structural changes associated with elevated BMI. Long-term monitoring is recommended to evaluate the progression and clinical impact of these findings. Full article

Effective and transparent medical diagnosis relies on accurate and interpretable classification of medical images across multiple modalities. This paper introduces an explainable multi-modal image analysis framework based on a dual-stream architecture that fuses handcrafted descriptors with deep features extracted from a custom MobileNet. Handcrafted descriptors include frequency-domain and texture features, while deep features are summarized using 26 statistical metrics to enhance interpretability. In the fusion stage, complementary features are combined at both the feature and decision levels. Decision-level integration combines calibrated soft voting, weighted voting, and stacking ensembles with optimized classifiers, including decision trees, random forests, gradient boosting, and logistic regression. To further refine performance, a hybrid class-specific feature selection strategy is proposed, combining mutual information, recursive elimination, and random forest importance to select the most discriminative features for each class. This hybrid selection approach eliminates redundancy, improves computational efficiency, and ensures robust classification. Explainability is provided through Local Interpretable Model-Agnostic Explanations, which offer transparent details about the ensemble model’s predictions and link influential handcrafted features to clinically meaningful image characteristics. The framework is validated on three benchmark datasets, i.e., BTTypes (brain MRI), Ultrasound Breast Images, and ACRIMA Retinal Fundus Images, demonstrating generalizability across modalities (MRI, ultrasound, retinal fundus) and disease categories (brain tumor, breast cancer, glaucoma). Full article

Background: Nuclear cataracts cause visual degradation through light scattering by aggregated proteins and particles within the crystalline lens. Existing computational models mainly consider submicron scatterers, while the optical impact of micrometer-scale particles observed in human nuclear cataracts remains underexplored. Objective: This study extends a LOCS III–based computational cataract model by incorporating micrometer-scale particles and quantitatively evaluates their effects on forward and backward light scattering across nuclear cataract grades. Methods: A physics-based scattering model was implemented using optical simulation software (LightTools). Three particle populations—nanometer-scale (S-type), submicron-scale (M-type), and micrometer-scale (L-type)—were uniformly distributed within the lens. Retinal luminance reduction was analyzed for forward scattering, while slit-lamp-based backward scattering simulations were used to evaluate luminance distributions and chromaticity changes. Particle concentrations were varied within clinically reported ranges corresponding to LOCS III grades. Results: Micrometer-scale particles had minimal impact in early nuclear cataract grades but significantly increased forward scattering and luminance loss in advanced grades (NO5–NO6). Backward scattering simulations revealed pronounced luminance enhancement and yellow chromaticity shifts with increasing micrometer-scale particle concentration. One micrometer-scale particle produced a luminance-reduction effect equivalent to approximately 6–7 submicron particles, depending on cataract severity. Conclusions: Including micrometer-scale particles enables a more complete optical representation of nuclear cataracts, linking retinal image degradation with slit-lamp appearance. The model provides a physically grounded framework for offline analysis and reference data generation to support clinical interpretation of cataract grading. Full article

Objectives: This study aimed to investigate the association between optical coherence tomography angiography (OCTA) parameters and cardiovascular (CV) risk scores in individuals with type 1 diabetes (T1D). Methods: A cross-sectional analysis of a large-scale prospective OCTA trial cohort (ClinicalTrials.gov NCT03422965) was performed. Demographic, systemic, and ocular data—including OCTA imaging—were collected. T1D participants were stratified into three CV risk categories: moderate (MR), high (HR), and very high risk (VHR). Individualized predictions for fatal and non-fatal CV events at 5 and 10 years were calculated using the STENO T1 Risk Engine calculator. Results: A total of 501 individuals (1 eye/patient; 397 T1D, 104 controls) were included. Subjects with MR (n = 37), HR (n = 152) and VHR (n = 208) exhibited significantly reduced vessel density (VD) (20.9 ± 1.3 vs. 20.2 ± 1.6 vs. 19.3 ± 1.8 mm⁻¹, p < 0.05), perfusion density (PD) (0.37 ± 0.02 vs. 0.36 ± 0.02 vs. 0.35 ± 0.02%, p < 0.05) and foveal avascular zone circularity (0.69 ± 0.06 vs. 0.65 ± 0.07 vs. 0.63 ± 0.09, p < 0.05). Statistically significant negative correlations were observed between CV risk and OCTA parameters including VD, PD, and retinal nerve fiber layer thickness, while central macular thickness (CMT) showed a positive correlation (p < 0.05). Notably, CMT was significantly associated with 5-year CV risk. Conclusions: OCTA-derived metrics, particularly reduced retinal VD and PD, are associated with elevated CV risk scores in T1D patients. These findings suggest that OCTA may serve as a valuable non-invasive tool for identifying individuals with increased CV risk scores. Full article