Medical History Taking and Differential Diagnosis on MIMIC-IV processed
46.1F1 ScoreNote2Chat-ST
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Note2Chat-STAverage Number of Turns=17.32026.01 | 46.1 | 46.2 | 54.5 | 70 | 81.2 | 84.4 | |
| Note2Chat-MTAverage Number of Turns=27.52026.01 | 43.8 | 55.4 | 41.8 | 62 | 78.2 | 82.6 | |
| DeepSeek-R1-0528-Qwen3-8BAverage Number of Turns=23.42026.01 | 29.6 | 34 | 32.7 | 37.2 | 51.6 | 61.2 | |
| GPT-4oAverage Number of Turns=22.92026.01 | 29.2 | 33.2 | 30.5 | 49 | 61.4 | 67.6 | |
| DoctorAgent-RLAverage Number of Turns=26.42026.01 | 28.4 | 35.1 | 27.5 | 35.6 | — | — | |
| MedGemma-27B-text-itAverage Number of Turns=21.42026.01 | 27.9 | 31.4 | 30.1 | 52.8 | 66.2 | 71.4 | |
| MedGemma-4B-itAverage Number of Turns=23.42026.01 | 27.2 | 31.6 | 28 | 40.6 | 55.2 | 62.2 | |
| Gemini-2.5-flashAverage Number of Turns=31.92026.01 | 26.6 | 35.5 | 26.7 | 51.4 | 66.2 | 73 | |
| o4-miniAverage Number of Turns=272026.01 | 23 | 28.7 | 21.9 | 47.6 | 60 | 67 | |
| Qwen2.5-7B-InstructAverage Number of Turns=10.32026.01 | 19.6 | 15.7 | 33 | 38.8 | 54.8 | 63.2 | |
| Qwen3-8BAverage Number of Turns=8.92026.01 | 17.9 | 13.8 | 34.1 | 33.4 | 46.6 | 55.2 | |
| HuatuoGPT-o1-8BAverage Number of Turns=2.022026.01 | 0.2 | 0.1 | 1.1 | 19.4 | 33 | 42.8 |