# Medical Report Summarization on MIMIC-CXR (test) Source: [Medical Report Summarization on MIMIC-CXR (test)]() Collection index: [llms.txt]() Data updated: 2026-04-01T08:49:01.183Z Evaluation of models summarizing medical reports using the test split of the MIMIC-CXR dataset, measured using BLEU-4, ROUGE-L, BERTScore, CheXbert, and RadGraph. 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 Report Summarization | | Dataset (including any stated split) | MIMIC-CXR (test) | | Metric | ROUGE-L | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 19 of 19 | | 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 - [ROUGE-L](): Higher is better; unit: PERCENT. - [BLEU-4](): Higher is better; unit: PERCENT. - [BERTScore](): Higher is better; unit: PERCENT. - [CheXbert](): Higher is better; unit: PERCENT. - [RadGraph](): Higher is better; unit: PERCENT. ## Recorded results | Method | ROUGE-L | Result date | Report paper | | --- | --- | --- | --- | | ViTAS | 69.83 | Not recorded | 69cca6a70636e5bf4afcbc9d | | Flan-T5 + ViT | 47.86 | Not recorded | 69cca6a70636e5bf4afcbc9d | | OFA + ResNet | 47.75 | Not recorded | 69cca6a70636e5bf4afcbc9d | | BLOOMZ-7b | 47.48 | Not recorded | 69cca6a70636e5bf4afcbc9d | | BART | 47 | Not recorded | 69cca6a70636e5bf4afcbc9d | | VG-BART (MHA) | 46.89 | Not recorded | 69cca6a70636e5bf4afcbc9d | | VG-BART (Dot) | 46.68 | Not recorded | 69cca6a70636e5bf4afcbc9d | | BioBART | 46.15 | Not recorded | 69cca6a70636e5bf4afcbc9d | | LLaVA-Med | 44.53 | Not recorded | 69cca6a70636e5bf4afcbc9d | | LLaVA-1.5 | 44.32 | Not recorded | 69cca6a70636e5bf4afcbc9d | | Qwen-VL | 44.3 | Not recorded | 69cca6a70636e5bf4afcbc9d | | BERT + CvT-21 | 44.14 | Not recorded | 69cca6a70636e5bf4afcbc9d | | DeepSeek-VL | 43.12 | Not recorded | 69cca6a70636e5bf4afcbc9d | | GSum | 42.98 | Not recorded | 69cca6a70636e5bf4afcbc9d | | WGSum | 41.55 | Not recorded | 69cca6a70636e5bf4afcbc9d | | PG | 36.84 | Not recorded | 69cca6a70636e5bf4afcbc9d | | Llama-3.1-8B | 32.46 | Not recorded | 69cca6a70636e5bf4afcbc9d | | Gemma-2-9b-it | 30.71 | Not recorded | 69cca6a70636e5bf4afcbc9d | | Mistral-7B | 29.81 | Not recorded | 69cca6a70636e5bf4afcbc9d | ### ViTAS ROUGE-L: 69.83. Result ID: 69ccdbea47f34ce924b6a399. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal"}. This record does not assert independent reproduction. ### Flan-T5 + ViT ROUGE-L: 47.86. Result ID: 69ccdbea47f34ce924b6a38e. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal"}. This record does not assert independent reproduction. ### OFA + ResNet ROUGE-L: 47.75. Result ID: 69ccdbea47f34ce924b6a38f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal"}. This record does not assert independent reproduction. ### BLOOMZ-7b ROUGE-L: 47.48. Result ID: 69ccdbea47f34ce924b6a390. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal"}. This record does not assert independent reproduction. ### BART ROUGE-L: 47. Result ID: 69ccdbea47f34ce924b6a388. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Text-Only"}. This record does not assert independent reproduction. ### VG-BART (MHA) ROUGE-L: 46.89. Result ID: 69ccdbea47f34ce924b6a393. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal","Attention Type":"Multi-Head Attention"}. This record does not assert independent reproduction. ### VG-BART (Dot) ROUGE-L: 46.68. Result ID: 69ccdbea47f34ce924b6a394. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal","Attention Type":"Dot Product"}. This record does not assert independent reproduction. ### BioBART ROUGE-L: 46.15. Result ID: 69ccdbea47f34ce924b6a391. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal"}. This record does not assert independent reproduction. ### LLaVA-Med ROUGE-L: 44.53. Result ID: 69ccdbea47f34ce924b6a398. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal"}. This record does not assert independent reproduction. ### LLaVA-1.5 ROUGE-L: 44.32. Result ID: 69ccdbea47f34ce924b6a395. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal"}. This record does not assert independent reproduction. ### Qwen-VL ROUGE-L: 44.3. Result ID: 69ccdbea47f34ce924b6a396. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal"}. This record does not assert independent reproduction. ### BERT + CvT-21 ROUGE-L: 44.14. Result ID: 69ccdbea47f34ce924b6a392. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal"}. This record does not assert independent reproduction. ### DeepSeek-VL ROUGE-L: 43.12. Result ID: 69ccdbea47f34ce924b6a397. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Multimodal"}. This record does not assert independent reproduction. ### GSum ROUGE-L: 42.98. Result ID: 69ccdbea47f34ce924b6a389. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Text-Only"}. This record does not assert independent reproduction. ### WGSum ROUGE-L: 41.55. Result ID: 69ccdbea47f34ce924b6a38a. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Text-Only"}. This record does not assert independent reproduction. ### PG ROUGE-L: 36.84. Result ID: 69ccdbea47f34ce924b6a387. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Text-Only"}. This record does not assert independent reproduction. ### Llama-3.1-8B ROUGE-L: 32.46. Result ID: 69ccdbea47f34ce924b6a38b. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Text-Only"}. This record does not assert independent reproduction. ### Gemma-2-9b-it ROUGE-L: 30.71. Result ID: 69ccdbea47f34ce924b6a38d. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Text-Only"}. This record does not assert independent reproduction. ### Mistral-7B ROUGE-L: 29.81. Result ID: 69ccdbea47f34ce924b6a38c. [Reporting paper]() Recorded evaluation configuration: {"Input Modality":"Text-Only"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.