# Medical Vision-Language Modeling on 3D-RAD Source: [Medical Vision-Language Modeling on 3D-RAD]() Collection index: [llms.txt]() Data updated: 2026-03-04T09:04:17.260Z Evaluation of Vision-Language Modeling performance on the 3D-RAD dataset, assessed using metrics such as BLEU-4, ROUGE-L, and LLM-Judge. 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 | Medical Vision-Language Modeling | | Dataset (including any stated split) | 3D-RAD | | Metric | BLEU-4 | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 12 of 12 | | 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 - [BLEU-4](): Higher is better; unit: SCALAR. - [ROUGE-L](): Higher is better; unit: SCALAR. - [B-Score](): Higher is better; unit: SCALAR. - [LLM-Judge Score](): Higher is better; unit: SCALAR. - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | BLEU-4 | Result date | Report paper | | --- | --- | --- | --- | | CT-Flow-8B | 36.96 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | CT-Flow-7B | 36.67 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | M3D-RAD | 29.76 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | Gemini-3-Pro-Preview | 29.59 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | GLM4.6-V | 23.23 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | GPT-5.2 | 22.08 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | Qwen3-VL-8B-Instruct | 20.89 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | Qwen3-VL-235B-A22B-Instruct | 20.55 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | Claude-Sonnet-4.5 | 20.14 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | Qwen2.5-VL-7B-Instruct | 18.33 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | Hulu-Med-7B | 12.77 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | | M3D-LaMed-Llama-2-7B | 12.33 | Not recorded | 69a66bdd2d603e6de3e3ccb2 | ### CT-Flow-8B BLEU-4: 36.96. Result ID: 69a7f4a96dbb7075b94cf3f4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tool-use":"true"}. This record does not assert independent reproduction. ### CT-Flow-7B BLEU-4: 36.67. Result ID: 69a7f4a96dbb7075b94cf3f3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tool-use":"true"}. This record does not assert independent reproduction. ### M3D-RAD BLEU-4: 29.76. Result ID: 69a7f4a96dbb7075b94cf3ef. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tool-use":"false"}. This record does not assert independent reproduction. ### Gemini-3-Pro-Preview BLEU-4: 29.59. Result ID: 69a7f4a96dbb7075b94cf3ea. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tool-use":"true"}. This record does not assert independent reproduction. ### GLM4.6-V BLEU-4: 23.23. Result ID: 69a7f4a96dbb7075b94cf3ed. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tool-use":"true"}. This record does not assert independent reproduction. ### GPT-5.2 BLEU-4: 22.08. Result ID: 69a7f4a96dbb7075b94cf3e9. [Reporting paper]() Recorded evaluation configuration: {"Tool-use":"true"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-Instruct BLEU-4: 20.89. Result ID: 69a7f4a96dbb7075b94cf3f2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tool-use":"true"}. This record does not assert independent reproduction. ### Qwen3-VL-235B-A22B-Instruct BLEU-4: 20.55. Result ID: 69a7f4a96dbb7075b94cf3ec. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tool-use":"true"}. This record does not assert independent reproduction. ### Claude-Sonnet-4.5 BLEU-4: 20.14. Result ID: 69a7f4a96dbb7075b94cf3eb. [Reporting paper]() Recorded evaluation configuration: {"Tool-use":"true"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B-Instruct BLEU-4: 18.33. Result ID: 69a7f4a96dbb7075b94cf3f1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tool-use":"true"}. This record does not assert independent reproduction. ### Hulu-Med-7B BLEU-4: 12.77. Result ID: 69a7f4a96dbb7075b94cf3f0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tool-use":"false"}. This record does not assert independent reproduction. ### M3D-LaMed-Llama-2-7B BLEU-4: 12.33. Result ID: 69a7f4a96dbb7075b94cf3ee. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tool-use":"false"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.