# Medical Visual Question Answering on M3D-Data (test)

Source: [Medical Visual Question Answering on M3D-Data (test)](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-m3d-data-test>)

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

Data updated: 2026-04-14T08:50:59.941Z

Evaluation of model performance on the test split of the M3D-Data dataset for Medical Visual Question Answering, reporting standard NLP metrics (BLEU, ROUGE, METEOR) alongside BERT and Accuracy scores.

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 | Medical Visual Question Answering |
| Dataset (including any stated split) | M3D-Data (test) |
| Metric | BLEU-1 |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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

- [BLEU-1](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-m3d-data-test.md?metric=BLEU-1>): Higher is better; unit: SCALAR.

- [ROUGE-1](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-m3d-data-test.md?metric=ROUGE-1>): Higher is better; unit: SCALAR.

- [METEOR](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-m3d-data-test.md?metric=METEOR>): Higher is better; unit: SCALAR.

- [BERT Score](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-m3d-data-test.md?metric=BERT%20Score>): Higher is better; unit: SCALAR.

- [Accuracy](<https://www.sota2.com/research/sota/medical-visual-question-answering-on-m3d-data-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | BLEU-1 | Result date | Report paper |
| --- | --- | --- | --- |
| TGH-MoE | 54.64 | Not recorded | 69ddca357d0172140ec79b53 |
| M3D | 52.1 | Not recorded | 69ddca357d0172140ec79b53 |
| Med3DVLM | 47.54 | Not recorded | 69ddca357d0172140ec79b53 |
| LLAVA-Med-1.5 | 34.53 | Not recorded | 69ddca357d0172140ec79b53 |
| RadFM | 16.39 | Not recorded | 69ddca357d0172140ec79b53 |

### TGH-MoE

BLEU-1: 54.64. Result ID: 69ddffa047f34ce924c582dc.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca357d0172140ec79b53.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ddca357d0172140ec79b53.md>)

Recorded evaluation configuration: {"Model Size":"4B","Trained on M3D-Data":"true"}. This record does not assert independent reproduction.

### M3D

BLEU-1: 52.1. Result ID: 69ddffa047f34ce924c582da.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca357d0172140ec79b53.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58507.md>)

Recorded evaluation configuration: {"Model Size":"4B","Trained on M3D-Data":"true"}. This record does not assert independent reproduction.

### Med3DVLM

BLEU-1: 47.54. Result ID: 69ddffa047f34ce924c582db.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca357d0172140ec79b53.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7aaf1e209131cea84981.md>)

Recorded evaluation configuration: {"Model Size":"7B","Trained on M3D-Data":"true"}. This record does not assert independent reproduction.

### LLAVA-Med-1.5

BLEU-1: 34.53. Result ID: 69ddffa047f34ce924c582d8.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca357d0172140ec79b53.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f8e1e209131cea46356.md>)

[Implementation](<https://aka.ms/llava-med>)

Recorded evaluation configuration: {"Model Size":"7B","Trained on M3D-Data":"false"}. This record does not assert independent reproduction.

### RadFM

BLEU-1: 16.39. Result ID: 69ddffa047f34ce924c582d9.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca357d0172140ec79b53.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71411e209131cea5023d.md>)

Recorded evaluation configuration: {"Model Size":"14B","Trained on M3D-Data":"false"}. This record does not assert independent reproduction.

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