Retinal image quality assessment using generic features
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Retinal image quality assessment is an important step in automated eye disease diagnosis. Diagnosis accuracy is highly dependent on the quality of retinal images, because poor image quality might prevent the observation of significant eye features and disease manifestations. A robust algorithm is therefore required in order to evaluate the quality of images in a large database. We developed an algorithm for retinal image quality assessment based on generic features that is independent from segmentation methods. It exploits the local sharpness and texture features by applying the cumulative probability of blur detection metric and run-length encoding algorithm, respectively. The quality features are combined to evaluate the image’s suitability for diagnosis purposes. Based on the recommendations of medical experts and our experience, we compared a global and a local approach. A support vector machine with radial basis functions was used as a nonlinear classifier in order to classify images to gradable and ungradable groups. We applied our methodology to 65 images of size 2592×1944 pixels that had been graded by a medical expert. The expert evaluated 38 images as gradable and 27 as ungradable. The results indicate very good agreement between the proposed algorithm’s predictions and the medical expert’s judgment: the sensitivity and specificity for the local approach are respectively 92% and 94%. The algorithm demonstrates sufficient robustness to identify relevant images for automated 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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it