Retinal Vessel Segmentation and Lesion Detection for Diabetic Retinopathy Diagnosis Using Context-Aware Deep Neural Networks on SPECTRAL OCT Data
Bibliographic record
Abstract
One of the primary causes of visual impairment in the world today is Diabetic Retinopathy (DR), which emphasizes the need for prompt and precise diagnosis to stop the progression of the illness.To improve retinal vascular segmentation and lesion identification, this work presents a unique architecture that combines Context Encoding Deep Neural Networks (CEDNN) with 3D-Spectral Optical Coherence Tomography (SPECTRAL OCT) imaging.The SPECTRAL OCT scans are subjected to Contrast Limited Adaptive Histogram Equalization (CLAHE) to enhance the image quality and for more accurate feature extraction.The CEDNN model utilizes context-encoding mechanisms to accurately segment retinal vessels by capturing both global and localized features.For lesion detection, specialized feature-encoding layers amplify subtle pathological signals while preserving structural details, correctly recognizing exudates, hemorrhages, and microaneurysms as DR indications.The CEDNN architecture enables accurate extraction of structural information for vessel segmentation, while the context-aware layers ensure reliable lesion identification without compromising spatial coherence.The model is validated on clinical 3D SPECTRAL OCT images and publicly available benchmark datasets, achieving outstanding performance with a 95% confidence interval for accuracy between 98.8% and 99.8%.The results show an F1-score of 99.3%, 99.4% sensitivity, 99% specificity, 99.2% precision, and 99.3% accuracy.These results highlight the model's robustness in capturing fine microvascular abnormalities and DR-related lesions across varied imaging conditions.The proposed 3D SPECTRAL OCT-based CEDNN framework offers a scalable, cost-effective solution for large-scale clinical screening and automated DR diagnosis.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".