Medical Report Generation on CT-RATE (test)
89.29BERTScore-F1CT2Rep
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| CT2Rep2026.06 | 89.29 | 32.12 | 45.43 | 58.99 | 52.47 | — | — | |
| RAD3D-PrefixLLM=DeepSeek-R1-Distill-LLaMA-8B, No. of Parameters in LLM=8B2026.06 | 89.09 | 34.05 | 44.59 | — | 54.43 | 47.04 | 43.39 | |
| RAD3D-Prefix2026.06 | 88.94 | 27.79 | 46.94 | 57.8 | 54.88 | — | — | |
| R2GenGPTLLM=LLaMA-2-7b-chat-hf, No. of Parameters in LLM=7B2026.06 | 88.86 | 35.23 | 45.09 | — | 50.41 | 46.4 | 40.38 | |
| RAD3D-PrefixLLM=LLaMA-3.2-1B, No. of Parameters in LLM=1B2026.06 | 88.83 | 36.37 | 46.94 | — | 54.88 | 42.56 | 41.9 | |
| E3D-GPT2026.06 | 87.97 | — | 41.79 | 52.6 | — | — | — | |
| R2GenGPTLLM=LLaMA-3.2-1B-Instruct, No. of Parameters in LLM=1B2026.06 | 87.51 | 29.02 | 37.62 | — | 41.2 | 39.485 | 34.68 | |
| CT-AGRG2026.06 | 86.7 | 17.2 | 19.6 | — | — | — | — | |
| R2GenGPTLLM=BioGPT-Large, No. of Parameters in LLM=1.6B2026.06 | 76.22 | 6.11 | 13.85 | — | 2.54 | 12.41 | 12.84 | |
| BaselineLLM=BioGPT-Large, No. of Parameters in LLM=1.6B2026.06 | 69.32 | — | — | — | — | — | — |