Open research questions in Retinal Imaging and Analysis
463 gap statements mined from Retinal Imaging and Analysis papers in our 7.9M-paper local library, which holds 2,469 papers on the topic — drawn mostly from each paper's own stated research gap, future-work, challenge and limitation notes, and its abstract. The ones listed below are a selection still marked open; each names the study that raised it, with a DOI link where the paper has one.
Choosing where to publish on Retinal Imaging and Analysis? See the ranked Medicine journals.
What the literature leaves open
The need to address class imbalance in the dataset. The challenge of providing integrated explainability and uncertainty quantification in the framework.
DAVis-Net: A Dual-Attention Deep Supervision Framework for Reliable Retinal OCT Image Classification with Integrated Explainability and Uncertainty Quantification · 2026 · DOILack of transparency in AI decision-making. Poor generalizability across diverse populations.
Why aren’t we using AI in eye clinics? A systematic review of barriers and solutions in AI-based fundus image diagnostics for ocular diseases · 2026 · DOIUsing an EfficientNet-B0 architecture, two models were trained on fundus images: one trained only on real images and another incorporating synthetic images generated by a Deep Convolutional Generative Adversarial Network (DCGAN) to simulate insufficient data.
Magnification correction with ML may improve characterisation of myopia-related microvascular changes and enhance diagnostic precision, warranting further study.
Correcting Axial Length‐Related Magnification Errors in Optical Coherence Tomography Angiography Improves Machine Learning Classification Performance in High Myopia · 2026 · DOIMoreover, an algorithm to detect glaucoma in highly myopic eyes has not yet been reported.
Detecting Glaucoma in Highly Myopic Eyes From Fundus Photographs Using Deep Convolutional Neural Networks · 2025 · DOIWhile data augmentation was applied to enhance model generalization, future studies could explore additional strategies such as focal loss or synthetic data generation to further strengthen robustness, particularly when extending to datasets that may contain imbalances or rare disease categories.
Evaluating the impact of preprocessing on CNN architectures: comparative analysis of a novel proposed model EyeDiagNet and existing models for eye disease detection · 2026 · DOIFuture research and development opportunities saturate when artificial intelligence (AI) is applied to the diagnosis of ocular disorders. Improving hybrid models which fuse deep learning and conventional machine learning remains a primary goal to achieve improved generalizability and performance.
Evaluating the impact of preprocessing on CNN architectures: comparative analysis of a novel proposed model EyeDiagNet and existing models for eye disease detection · 2026 · DOIThe need for systematic screening programs for retinopathy detection in people with diabetes. The lack of studies evaluating the gradable rate of retinal images acquired with DRSplus retinographer in patients with diabetes.
Future research should focus on developing AI systems that can detect ophthalmological diseases at an early stage. Future research should investigate the use of AI in ophthalmology in different populations and settings. Future research should address the ethical concerns related to the use of AI in healthcare.
APPLICATION OF ARTIFICIAL INTELLIGENCE IN OPHTHALMOLOGY AND MEDICAL ETHICS PRINCIPLES (LITERATURE REVIEW) · 2026 · DOIThere is a need for more research on the use of AI in ophthalmology. The development of AI systems for ophthalmology is limited by the availability of high-quality medical images. There are ethical concerns related to the use of AI in healthcare.
APPLICATION OF ARTIFICIAL INTELLIGENCE IN OPHTHALMOLOGY AND MEDICAL ETHICS PRINCIPLES (LITERATURE REVIEW) · 2026 · DOILimited portability across systems. Hard-coded file paths. Missing scripts or modules. Dependency conflicts and outdated software libraries.
Evaluating the real-world accessibility of open-source deep learning models for retinal vessel segmentation in optical coherence tomography angiography · 2026 · DOIThe dataset was limited to 182 FFA images. The method requires expert-annotated labels. The method may not generalize to other datasets.
An intelligent segmentation method for leakage points in central serous chorioretinopathy based on fluorescein angiography images · 2026 · DOIEnhanced data collection including detailed patient demographics and clinical characteristics will enable more comprehensive performance evaluation. Large-scale multi-center validation studies will improve algorithm generalizability across diverse populations.
Diagnostic Performance of Artificial Intelligence and Deep Learning for Diabetic Retinopathy Screening: A Systematic Review and Meta-analysis · 2026 · DOIThe integration of artificial intelligence (AI) with advanced ocular drug delivery systems is expected to significantly transform ophthalmic therapeutics. With increasing computational power and availability of large clinical datasets, AI models are likely to evolve from supportive analytical tools to integrated systems that assist therapeutic optimization.
ARTIFICIAL INTELLIGENCE (AI) ENABLED PERSONALIZED MEDICINE - SHAPING THE FUTURE OF OCULAR THERAPEUTICS · 2026 · DOIMulti-center validation of the DAVis-Net model, - Clinical validation of the Trust/Refer triage rule, - Further research on the application of DAVis-Net to other medical image classification…
DAVis-Net: A Dual-Attention Deep Supervision Framework for Reliable Retinal OCT Image Classification with Integrated Explainability and Uncertainty Quantification · 2026 · DOIInter-annotator consistency among junior ophthalmologists - Discrepancies in vessel identification, boundary delineation, vascular continuity, or class assignment - Limited availability of publicly accessible pediatric retinal imaging…
Limited availability of publicly accessible pediatric retinal imaging datasets - Constraint on large-scale retinal vascular research and development of related algorithms - Need for a dataset focused on pediatric…
The proposed framework is primarily designed for scenarios where multimodal imaging data are accessible, and its applicability may be limited in situations where such data are not routinely collected - Further validation is required to establish their advantage over unimodal approaches - The fusion mechanisms explored in this study are widely used conventional fusion…
Generating the necessary modality images through image synthesis - Refining fusion strategies - Expanding this approach to broader clinical applications for improved real-world implementation - Investigating more advanced cross-modal fusion strategies, such as attention-based and transformer-based…
Exploring other potential applications of PCRNet, - Investigating the effectiveness of PCRNet on other medical image analysis tasks, - Developing more advanced loss functions to further improve the performance…
The existing deep neural networks (DNNs) often ignore exploring the clinical pathology context and expert experience priors. The classical cross-entropy (CE) loss neglects class imbalance, degenerating the performance of classes with few examples.
The study only evaluated the top ten performing models, - The study did not assess the models' performance on a larger dataset, - The study did not compare the models' performance with other state-of-the-art methods, - The study was limited by the availability of computational…
Evaluating the real-world accessibility of open-source deep learning models for retinal vessel segmentation in optical coherence tomography angiography · 2026 · DOIUnderreported metrics, fractal dimension, parafoveal rim VLD and BD, emerged as key discriminators.
Correcting Axial Length‐Related Magnification Errors in Optical Coherence Tomography Angiography Improves Machine Learning Classification Performance in High Myopia · 2026 · DOIThe quantitative comparison was conducted on only 36 images out of 40 test FFA images where leakage points were successfully detected, limiting the generalizability…
An intelligent segmentation method for leakage points in central serous chorioretinopathy based on fluorescein angiography images · 2026 · DOIDeveloping efficient, reliable, and clinically adaptable systems suitable for real-world deployment. Exploring optimization techniques such as model pruning and quantization. Investigating privacy-preserving collaborative learning approaches, including decentralized or federated learning frameworks.
AI-Driven Automated Screening of Diabetic Retinopathy: A Comprehensive Review of Convolutional Neural Networks for Early Detection · 2026 · DOI
Most-cited papers in Retinal Imaging and Analysis
- Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs · JAMA · 2016 · 6,699 citations
- Development and Validation of a Deep Learning System for Diabetic Retinopathy and Related Eye Diseases Using Retinal Images From Multiethnic Populations With Diabetes · JAMA · 2017 · 2,175 citations
- Retinal Vascular Layers Imaged by Fluorescein Angiography and Optical Coherence Tomography Angiography · JAMA Ophthalmology · 2015 · 1,392 citations
- Unified Focal loss: Generalising Dice and cross entropy-based losses to handle class imbalanced medical image segmentation · Computerized Medical Imaging and Graphics · 2021 · 716 citations
- A deep learning system for detecting diabetic retinopathy across the disease spectrum · Nature Communications · 2021 · 566 citations
- Diabetic retinopathy detection through deep learning techniques: A review · Informatics in Medicine Unlocked · 2020 · 455 citations
- CABNet: Category Attention Block for Imbalanced Diabetic Retinopathy Grading · IEEE Transactions on Medical Imaging · 2020 · 376 citations
- Applications of deep learning in fundus images: A review · Medical Image Analysis · 2021 · 371 citations
- Uncertainty quantification using Bayesian neural networks in classification: Application to biomedical image segmentation · Computational Statistics & Data Analysis · 2019 · 370 citations
- Artificial Intelligence to Detect Papilledema from Ocular Fundus Photographs · New England Journal of Medicine · 2020 · 352 citations
Most recent work
- Difficulty-Aware Pseudo-Label Correction Network for Fine-Grained Classification of Choroidal Neovascularization in OCT Images · IEEE Transactions on Multimedia · 2026
- AI framework for multidisease detection via retinal imaging · Nature Medicine · 2026
- Enhanced Diabetic Retinopathy Detection: An Explainable Semi-Supervised Approach Using Contrastive Learning · IEEE Journal of Biomedical and Health Informatics · 2026
- Explainable Multi-Modal Medical Image Analysis Through Dual-Stream Multi-Feature Fusion and Class-Specific Selection · AI · 2026
- Artificial intelligence based retinal imaging for cardiovascular risk and statin guidance in retinal vein occlusion · American Journal of Preventive Cardiology · 2026
- HOG-CNN: Integrating Histogram of Oriented Gradients with Convolutional Neural Networks for Retinal Image Classification · Journal of Imaging Informatics in Medicine · 2026
- M2Net: Multimodal Multitask Mutual Learning for Anti-VEGF Efficacy Prediction · IEEE Transactions on Medical Imaging · 2026
- Enhancing ocular sign detection: AI-based strategic segmentation for improved accuracy and privacy protection · npj Digital Medicine · 2026
- Artificial Intelligence-Based Medical Devices for Diabetic Retinopathy Screening in the European Union · Ophthalmology and Therapy · 2026
- Structure-preserving super-resolution of retinal fundus images via a dual-transformer residual network · Frontiers in Medicine · 2026
Find a gap in your own Retinal Imaging and Analysis sub-topic
This page shows what the Retinal Imaging and Analysis literature already flags as unresolved. To narrow it to your specific question, search the Research Gap Finder: the search is free with a free account and lists the papers closest to your topic first. Unlocking that topic (50 credits, charged once) fills the comparison table from our 7.9M-paper local library and writes the gaps from its rows.
Open the Research Gap Finder →