A stochastically fully connected conditional random field framework for super resolution OCT
Bibliographic record
Abstract
A number of factors can degrade the resolution and contrast of OCT images, such as: (1) changes of the OCT pointspread function (PSF) resulting from wavelength dependent scattering and absorption of light along the imaging depth (2) speckle noise, as well as (3) motion artifacts. We propose a new Super Resolution OCT (SR OCT) imaging framework that takes advantage of a Stochastically Fully Connected Conditional Random Field (SF-CRF) model to generate a Super Resolved OCT (SR OCT) image of higher quality from a set of Low-Resolution OCT (LR OCT) images. The proposed SF-CRF SR OCT imaging is able to simultaneously compensate for all of the factors mentioned above, that degrade the OCT image quality, using a unified computational framework. The proposed SF-CRF SR OCT imaging framework was tested on a set of simulated LR human retinal OCT images generated from a high resolution, high contrast retinal image, and on a set of in-vivo, high resolution, high contrast rat retinal OCT images. The reconstructed SR OCT images show considerably higher spatial resolution, less speckle noise and higher contrast compared to other tested methods. Visual assessment of the results demonstrated the usefulness of the proposed approach in better preservation of fine details and structures of the imaged sample, retaining biological tissue boundaries while reducing speckle noise using a unified computational framework. Quantitative evaluation using both Contrast to Noise Ratio (CNR) and Edge Preservation (EP) parameter also showed superior performance of the proposed SF-CRF SR OCT approach compared to other image processing approaches.
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 | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| 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".