# OCT Classification on Seven Downstream OCT Classification Datasets (test)

Source: [OCT Classification on Seven Downstream OCT Classification Datasets (test)](<https://www.sota2.com/research/sota/oct-classification-on-seven-downstream-oct-classification-datasets-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:36:36.611Z

Evaluation of Optical Coherence Tomography (OCT) classification performance across a suite of seven specific downstream OCT datasets, measured using Balanced Accuracy (BAcc), Area Under the ROC Curve (AUROC), and Average Precision (AP).

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | OCT classification |
| Dataset (including any stated split) | Seven downstream OCT classification datasets (test) |
| Metric | BAcc |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [BAcc](<https://www.sota2.com/research/sota/oct-classification-on-seven-downstream-oct-classification-datasets-test.md?metric=BAcc>): Higher is better; unit: PERCENT.

- [AUROC](<https://www.sota2.com/research/sota/oct-classification-on-seven-downstream-oct-classification-datasets-test.md?metric=AUROC>): Higher is better; unit: PERCENT.

- [AP](<https://www.sota2.com/research/sota/oct-classification-on-seven-downstream-oct-classification-datasets-test.md?metric=AP>): Higher is better; unit: PERCENT.

## Recorded results

| Method | BAcc | Result date | Report paper |
| --- | --- | --- | --- |
| RetFiner-R | 83.8 | Not recorded | 68fa79b51e209131cea7f063 |
| RetFiner-V | 83.5 | Not recorded | 68fa79b51e209131cea7f063 |
| RetFiner-U | 82.5 | Not recorded | 68fa79b51e209131cea7f063 |
| VisionFM | 81.4 | Not recorded | 68fa79b51e209131cea7f063 |
| UrFound | 78.6 | Not recorded | 68fa79b51e209131cea7f063 |
| RETFound | 78.1 | Not recorded | 68fa79b51e209131cea7f063 |
| DINOv2 | 74.5 | Not recorded | 68fa79b51e209131cea7f063 |
| CLIP | 73.6 | Not recorded | 68fa79b51e209131cea7f063 |

### RetFiner-R

BAcc: 83.8. Result ID: 698b5a687a967eddb76ddf30.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

[Implementation](<https://github.com/ronnief1/RetFiner>)

Recorded evaluation configuration: {"Protocol":"Linear probing","Base Model":"RETFound"}. This record does not assert independent reproduction.

### RetFiner-V

BAcc: 83.5. Result ID: 698b5a687a967eddb76ddf32.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

[Implementation](<https://github.com/ronnief1/RetFiner>)

Recorded evaluation configuration: {"Protocol":"Linear probing","Base Model":"VisionFM"}. This record does not assert independent reproduction.

### RetFiner-U

BAcc: 82.5. Result ID: 698b5a687a967eddb76ddf31.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

[Implementation](<https://github.com/ronnief1/RetFiner>)

Recorded evaluation configuration: {"Protocol":"Linear probing","Base Model":"UrFound"}. This record does not assert independent reproduction.

### VisionFM

BAcc: 81.4. Result ID: 698b5a687a967eddb76ddf2f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

Recorded evaluation configuration: {"Protocol":"Linear probing"}. This record does not assert independent reproduction.

### UrFound

BAcc: 78.6. Result ID: 698b5a687a967eddb76ddf2e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74721e209131cea61d9f.md>)

Recorded evaluation configuration: {"Protocol":"Linear probing"}. This record does not assert independent reproduction.

### RETFound

BAcc: 78.1. Result ID: 698b5a687a967eddb76ddf2d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

Recorded evaluation configuration: {"Protocol":"Linear probing"}. This record does not assert independent reproduction.

### DINOv2

BAcc: 74.5. Result ID: 698b5a687a967eddb76ddf2c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72181e209131cea546c3.md>)

[Implementation](<https://github.com/facebookresearch/dinov2>)

Recorded evaluation configuration: {"Protocol":"Linear probing"}. This record does not assert independent reproduction.

### CLIP

BAcc: 73.6. Result ID: 698b5a687a967eddb76ddf2b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79b51e209131cea7f063.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"Protocol":"Linear probing"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.