Deep Learning-Based Segmentation of Geographic Atrophy: A Multi-Center, Multi-Device Validation in a Real-World Clinical Cohort
Bibliographic record
Abstract
Background: To report a deep learning-based algorithm for automated segmentation of geographic atrophy (GA) among patients with age-related macular degeneration (AMD). Methods: Validation of a deep learning algorithm was performed using optical coherence tomography (OCT) images from patients in routine clinical care diagnosed with GA, with and without concurrent nAMD. For model construction, a 3D U-Net architecture was used with the output modified to generate a 2D mask. Accuracy of the model was assessed relative to the manual labeling of GA with the Dice similarity coefficient (DSC) and correlation r2 scores. Results: The OCT data set included 367 scans from the Spectralis (Heidelberg, Germany) from 55 eyes in 33 subjects; 267 (73%) scans had concurrent nAMD. In parallel, 348 scans were collected using the Cirrus (Zeiss), from 348 eyes in 326 subjects; 101 (29%) scans had concurrent nAMD. For Spectralis data, the mean DSC score was 0.83 and r2 was 0.91. For Cirrus data, the mean DSC score was 0.82 and r2 was 0.88. Conclusions: The reported deep learning algorithm demonstrated strong agreement with manual grading of GA secondary to AMD on the OCT data set from routine clinical practice. The model performed well across two OCT devices as well as amongst patients with GA with concurrent nAMD, suggesting applicability in the clinical space.
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 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 | 0.000 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, 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".