# Lesion Segmentation on ComLesion-14K 1.0 (test)

Source: [Lesion Segmentation on ComLesion-14K 1.0 (test)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test>)

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

Data updated: 2026-03-09T07:49:36.328Z

Evaluation of lesion segmentation performance on the test split of the ComLesion-14K 1.0 dataset, reporting mean Dice (mDice), mean IoU (mIoU), Failure Rate, and Dice scores for specific anatomical regions (Head, Abd., Lung, Eye, Neck, Skin, Foot, Breast, Heart).

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 | Lesion Segmentation |
| Dataset (including any stated split) | ComLesion-14K 1.0 (test) |
| Metric | Mean Dice (mDice) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 14 of 14 |
| 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

- [Mean Dice (mDice)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Mean%20Dice%20(mDice)>): Higher is better; unit: PERCENT.

- [Mean IoU (mIoU)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Mean%20IoU%20(mIoU)>): Higher is better; unit: PERCENT.

- [Failure Rate](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Failure%20Rate>): Lower is better; unit: PERCENT.

- [Dice (Head)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Dice%20(Head)>): Higher is better; unit: PERCENT.

- [Dice (Abdomen)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Dice%20(Abdomen)>): Higher is better; unit: PERCENT.

- [Dice (Lung)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Dice%20(Lung)>): Higher is better; unit: PERCENT.

- [Dice (Eye)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Dice%20(Eye)>): Higher is better; unit: PERCENT.

- [Dice (Neck)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Dice%20(Neck)>): Higher is better; unit: PERCENT.

- [Dice (Skin)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Dice%20(Skin)>): Higher is better; unit: PERCENT.

- [Dice (Foot)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Dice%20(Foot)>): Higher is better; unit: PERCENT.

- [Dice (Breast)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Dice%20(Breast)>): Higher is better; unit: PERCENT.

- [Dice (Heart)](<https://www.sota2.com/research/sota/lesion-segmentation-on-comlesion-14k-1-0-test.md?metric=Dice%20(Heart)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Mean Dice (mDice) | Result date | Report paper |
| --- | --- | --- | --- |
| CORE-Seg | 37.06 | Not recorded | 69ae53d82d603e6de3e42801 |
| LISA-3B (Trained) | 22.17 | Not recorded | 69ae53d82d603e6de3e42801 |
| SegZero-3B (Trained) | 17.47 | Not recorded | 69ae53d82d603e6de3e42801 |
| MedPLIB-7B | 12.9 | Not recorded | 69ae53d82d603e6de3e42801 |
| MediSee-7B | 11.07 | Not recorded | 69ae53d82d603e6de3e42801 |
| Qwen2.5-VL-72B | 11.04 | Not recorded | 69ae53d82d603e6de3e42801 |
| SegZero-7B | 9.08 | Not recorded | 69ae53d82d603e6de3e42801 |
| Qwen2.5-VL-7B | 8.3 | Not recorded | 69ae53d82d603e6de3e42801 |
| Lingshu-7B | 7.37 | Not recorded | 69ae53d82d603e6de3e42801 |
| MedGemma-4B | 6.03 | Not recorded | 69ae53d82d603e6de3e42801 |
| VLMR1-REC-3B | 4.76 | Not recorded | 69ae53d82d603e6de3e42801 |
| HuatuoGPT-7B | 4.73 | Not recorded | 69ae53d82d603e6de3e42801 |
| InternVL3-8B | 4.18 | Not recorded | 69ae53d82d603e6de3e42801 |
| Chiron-o1-8B | 3.35 | Not recorded | 69ae53d82d603e6de3e42801 |

### CORE-Seg

Mean Dice (mDice): 37.06. Result ID: 69ae7afc6dbb7075b956576b.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

Recorded evaluation configuration: {"Method Category":"Grounding-Specific MLLMs","Evaluation Protocol":"Trained"}. This record does not assert independent reproduction.

### LISA-3B (Trained)

Mean Dice (mDice): 22.17. Result ID: 69ae7afc6dbb7075b9565769.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa730a1e209131cea5a0b2.md>)

[Implementation](<https://www.sota2.com/github.com/dvlab-research/LISA>)

Recorded evaluation configuration: {"Method Category":"Grounding-Specific MLLMs","Evaluation Protocol":"Trained"}. This record does not assert independent reproduction.

### SegZero-3B (Trained)

Mean Dice (mDice): 17.47. Result ID: 69ae7afc6dbb7075b956576a.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79b81e209131cea7f19c.md>)

Recorded evaluation configuration: {"Method Category":"Grounding-Specific MLLMs","Evaluation Protocol":"Trained"}. This record does not assert independent reproduction.

### MedPLIB-7B

Mean Dice (mDice): 12.9. Result ID: 69ae7afc6dbb7075b9565767.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bf21e209131cea8b849.md>)

Recorded evaluation configuration: {"Method Category":"Grounding-Specific MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

### MediSee-7B

Mean Dice (mDice): 11.07. Result ID: 69ae7afc6dbb7075b9565768.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa783c1e209131cea775ad.md>)

Recorded evaluation configuration: {"Method Category":"Grounding-Specific MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

### Qwen2.5-VL-72B

Mean Dice (mDice): 11.04. Result ID: 69ae7afc6dbb7075b956575e.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"Method Category":"General MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

### SegZero-7B

Mean Dice (mDice): 9.08. Result ID: 69ae7afc6dbb7075b9565765.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79b81e209131cea7f19c.md>)

Recorded evaluation configuration: {"Method Category":"Grounding-Specific MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

### Qwen2.5-VL-7B

Mean Dice (mDice): 8.3. Result ID: 69ae7afc6dbb7075b9565760.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"Method Category":"General MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

### Lingshu-7B

Mean Dice (mDice): 7.37. Result ID: 69ae7afc6dbb7075b9565763.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa795a1e209131cea7d387.md>)

Recorded evaluation configuration: {"Method Category":"Medical-Specific MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

### MedGemma-4B

Mean Dice (mDice): 6.03. Result ID: 69ae7afc6dbb7075b9565764.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79ff1e209131cea808a4.md>)

Recorded evaluation configuration: {"Method Category":"Medical-Specific MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

### VLMR1-REC-3B

Mean Dice (mDice): 4.76. Result ID: 69ae7afc6dbb7075b9565766.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa780b1e209131cea7681f.md>)

Recorded evaluation configuration: {"Method Category":"Grounding-Specific MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

### HuatuoGPT-7B

Mean Dice (mDice): 4.73. Result ID: 69ae7afc6dbb7075b9565761.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

Recorded evaluation configuration: {"Method Category":"Medical-Specific MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

### InternVL3-8B

Mean Dice (mDice): 4.18. Result ID: 69ae7afc6dbb7075b956575f.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78241e209131cea76fa5.md>)

Recorded evaluation configuration: {"Method Category":"General MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

### Chiron-o1-8B

Mean Dice (mDice): 3.35. Result ID: 69ae7afc6dbb7075b9565762.

[Reporting paper](<https://www.sota2.com/research/paper/69ae53d82d603e6de3e42801.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c701e209131cea8e14d.md>)

Recorded evaluation configuration: {"Method Category":"Medical-Specific MLLMs","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction.

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