StanfordAIMI/GREEN-V2

--- license: apache-2.0 task_categories: - text-generation language: - en tags: - medical size_categories: - 100K<n<1M --- # GREEN Dataset We share the dataset used to train the LLM metric introduced in ["GREEN: Generative Radiology Report Evaluation and Error Notation"](https://arxiv.org/pdf/2405.03595). GREEN is a evaluation metric for radiology reports that uses language models to identify and explain clinically significant errors, offering better alignment with expert preferences and more interpretable results compared to existing metrics. The method provides both quantitative scores and qualitative explanations, has been validated against expert assessments and GPT-4, and offers an open-source alternative that performs comparably to commercial solutions. ## Dataset Details For details, please check the paper, [project page](https://stanford-aimi.github.io/green.html), [Github](https://github.com/Stanford-AIMI/GREEN) ### Dataset Description - **Curated by:** [JB Delbrouk, Zhihong Chen, Sophie Ostmeier] - **Funded by [optional]:** Stanford AIMI Center, German Research Foundation - **Language(s) (NLP):** English - **License:** Apache license 2.0 - ### Out-of-Scope Use The dataset includes 50k examples that are not Chest Xray reports. However, currently the model and the dataset are primarily designed to be used for Chest Xrays. ## Citation <!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. --> **BibTeX:** ``` @article{ostmeier2024green, title={GREEN: Generative Radiology Report Evaluation and Error Notation}, author={Ostmeier, Sophie and Xu, Justin and Chen, Zhihong and Varma, Maya and Blankemeier, Louis and Bluethgen, Christian and Michalson, Arne Edward and Moseley, Michael and Langlotz, Curtis and Chaudhari, Akshay S and others}, journal={arXiv preprint arXiv:2405.03595}, year={2024} } ```许可证：Apache 2.0 任务类别： - 文本生成 语言： - 英语 标签： - 医疗 样本量区间： - 10万<样本量<100万 # GREEN数据集 我们共享了发表于论文《GREEN: 生成式放射科报告评估与错误标注》（arXiv:2405.03595）中提出的用于训练大语言模型（LLM）评估指标的数据集。 GREEN是一款面向放射科报告的评估指标，借助语言模型识别并解释具有临床意义的错误。相较于现有评估指标，其与专家偏好的对齐度更高，结果也更具可解释性。该方法可同时输出量化评分与质性解释，已通过专家评估与GPT-4完成验证，同时提供了可与商用方案性能媲美的开源替代方案。 ## 数据集详情 详细信息请参阅原论文、[项目页面](https://stanford-aimi.github.io/green.html)及[GitHub仓库](https://github.com/Stanford-AIMI/GREEN)。 ### 数据集描述 - **整理者：** [JB·德尔布鲁克（JB Delbrouk）、陈志宏（Zhihong Chen）、索菲·奥斯特迈尔（Sophie Ostmeier）] - **资助方（可选）：** 斯坦福AI医学影像中心（Stanford AIMI Center）、德国研究基金会（German Research Foundation） - **自然语言处理所用语言：** 英语 - **许可证：** Apache 2.0许可证 ### 适用范围限制 本数据集包含5万个非胸部X线报告的样本，但当前模型与数据集主要面向胸部X线报告场景设计。 ## 引用 > 若有介绍该数据集的论文或博客文章，需在此处附上APA及BibTeX格式的引用信息。 **BibTeX：** @article{ostmeier2024green, title={GREEN: Generative Radiology Report Evaluation and Error Notation}, author={Ostmeier, Sophie and Xu, Justin and Chen, Zhihong and Varma, Maya and Blankemeier, Louis and Bluethgen, Christian and Michalson, Arne Edward and Moseley, Michael and Langlotz, Curtis and Chaudhari, Akshay S and others}, journal={arXiv preprint arXiv:2405.03595}, year={2024} }