# X-ray understanding on MIMIC-CXR uncertain as negative

Source: [X-ray understanding on MIMIC-CXR uncertain as negative](<https://www.sota2.com/research/sota/x-ray-understanding-on-mimic-cxr-uncertain-as-negative>)

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

Data updated: 2026-02-26T05:52:44.196Z

Evaluation of general X-ray understanding capabilities on the specific subset of the MIMIC-CXR dataset where findings were marked as uncertain or negative, measured using Micro and Macro F1 scores across 5 and 14 categories.

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 | X-ray understanding |
| Dataset (including any stated split) | MIMIC-CXR uncertain as negative |
| Metric | Micro F1 (14 classes) |
| 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

- [Micro F1 (14 classes)](<https://www.sota2.com/research/sota/x-ray-understanding-on-mimic-cxr-uncertain-as-negative.md?metric=Micro%20F1%20(14%20classes)>): Higher is better; unit: PERCENT.

- [Micro F1 (5 classes)](<https://www.sota2.com/research/sota/x-ray-understanding-on-mimic-cxr-uncertain-as-negative.md?metric=Micro%20F1%20(5%20classes)>): Higher is better; unit: PERCENT.

- [Macro F1 (14 classes)](<https://www.sota2.com/research/sota/x-ray-understanding-on-mimic-cxr-uncertain-as-negative.md?metric=Macro%20F1%20(14%20classes)>): Higher is better; unit: PERCENT.

- [Macro F1 (5 classes)](<https://www.sota2.com/research/sota/x-ray-understanding-on-mimic-cxr-uncertain-as-negative.md?metric=Macro%20F1%20(5%20classes)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Micro F1 (14 classes) | Result date | Report paper |
| --- | --- | --- | --- |
| LLaVA-Rad | 57.3 | Not recorded | 696dba8ee25ad2fa9d1ecd8c |
| Med-PaLM M | 53.6 | Not recorded | 696dba8ee25ad2fa9d1ecd8c |
| UniX | 53.6 | Not recorded | 696dba8ee25ad2fa9d1ecd8c |
| RGRG | 38.9 | Not recorded | 696dba8ee25ad2fa9d1ecd8c |
| GPT-4V | 35.5 | Not recorded | 696dba8ee25ad2fa9d1ecd8c |
| LLaVA-Med | 27.2 | Not recorded | 696dba8ee25ad2fa9d1ecd8c |
| HealthGPT | 24.2 | Not recorded | 696dba8ee25ad2fa9d1ecd8c |
| PromptMRG | 15.3 | Not recorded | 696dba8ee25ad2fa9d1ecd8c |

### LLaVA-Rad

Micro F1 (14 classes): 57.3. Result ID: 6997b3605698665ec4098a79.

[Reporting paper](<https://www.sota2.com/research/paper/696dba8ee25ad2fa9d1ecd8c.md>)

Recorded evaluation configuration: {"Und. Params":"7B"}. This record does not assert independent reproduction.

### Med-PaLM M

Micro F1 (14 classes): 53.6. Result ID: 6997b3605698665ec4098a78.

[Reporting paper](<https://www.sota2.com/research/paper/696dba8ee25ad2fa9d1ecd8c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70281e209131cea497c0.md>)

Recorded evaluation configuration: {"Und. Params":"84B"}. This record does not assert independent reproduction.

### UniX

Micro F1 (14 classes): 53.6. Result ID: 6997b3605698665ec4098a7e.

[Reporting paper](<https://www.sota2.com/research/paper/696dba8ee25ad2fa9d1ecd8c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/696dba8ee25ad2fa9d1ecd8c.md>)

[Implementation](<https://github.com/ZrH42/UniX>)

Recorded evaluation configuration: {"Und. Params":"1.5B"}. This record does not assert independent reproduction.

### RGRG

Micro F1 (14 classes): 38.9. Result ID: 6997b3605698665ec4098a7c.

[Reporting paper](<https://www.sota2.com/research/paper/696dba8ee25ad2fa9d1ecd8c.md>)

Recorded evaluation configuration: {"Und. Params":"\<1B"}. This record does not assert independent reproduction.

### GPT-4V

Micro F1 (14 classes): 35.5. Result ID: 6997b3605698665ec4098a77.

[Reporting paper](<https://www.sota2.com/research/paper/696dba8ee25ad2fa9d1ecd8c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70d71e209131cea4d979.md>)

Recorded evaluation configuration: {"Und. Params":null}. This record does not assert independent reproduction.

### LLaVA-Med

Micro F1 (14 classes): 27.2. Result ID: 6997b3605698665ec4098a7a.

[Reporting paper](<https://www.sota2.com/research/paper/696dba8ee25ad2fa9d1ecd8c.md>)

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

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

Recorded evaluation configuration: {"Und. Params":"7B"}. This record does not assert independent reproduction.

### HealthGPT

Micro F1 (14 classes): 24.2. Result ID: 6997b3605698665ec4098a7d.

[Reporting paper](<https://www.sota2.com/research/paper/696dba8ee25ad2fa9d1ecd8c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77201e209131cea713ef.md>)

Recorded evaluation configuration: {"Und. Params":"3.8B"}. This record does not assert independent reproduction.

### PromptMRG

Micro F1 (14 classes): 15.3. Result ID: 6997b3605698665ec4098a7b.

[Reporting paper](<https://www.sota2.com/research/paper/696dba8ee25ad2fa9d1ecd8c.md>)

Recorded evaluation configuration: {"Und. Params":"\<1B"}. This record does not assert independent reproduction.

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