Measurement on DeepTumorVQA refined 2025b
42Lesion Volume Measurement3DMedAgent (GPT-5)
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| 3DMedAgent (GPT-5)Base Model=GPT-5, Evaluation Protocol=zero-shot2026.02 | 42 | 82 | 63 | 63 | |
| 3DMedAgent (Qwen3-VL)Base Model=Qwen3-VL (30B), Evaluation Protocol=zero-shot2026.02 | 35 | 78 | 63 | 59 | |
| Qwen3-VLParameters=30B, Evaluation Protocol=zero-shot2026.02 | 30 | 35 | 41 | 35 | |
| RandType=Random baseline2026.02 | 25 | 25 | 25 | 25 | |
| GPT-5Evaluation Protocol=zero-shot2026.02 | 25 | 35 | 39 | 33 | |
| RadFMEvaluation Protocol=zero-shot2026.02 | 25 | 37 | 31 | 31 | |
| MedGemmaParameters=27B, Evaluation Protocol=zero-shot2026.02 | 24 | 21 | 41 | 29 | |
| HuatuoGPTParameters=34B, Evaluation Protocol=zero-shot2026.02 | 20 | 35 | 24 | 26 | |
| M3DEvaluation Protocol=zero-shot2026.02 | 18 | 33 | 32 | 28 |