LCD benchmark: long clinical document benchmark on mortality prediction for language models.
LCD 基準:針對語言模型的死亡預測長期臨床文件基準。
J Am Med Inform Assoc 2024-11-27
Benchmarking large language models for biomedical natural language processing applications and recommendations.
大型語言模型在生物醫學自然語言處理應用中的基準測試與建議。
Nat Commun 2025-04-05
Comparative benchmarking of the DeepSeek large language model on medical tasks and clinical reasoning.
DeepSeek 大型語言模型在醫學任務與臨床推理上的比較性基準評估
Nat Med 2025-04-23
Use of Open-Source Large Language Models for Automatic Synthesis of the Entire Imaging Medical Records of Patients: A Feasibility Study.
開源大型語言模型於自動整合病患完整影像醫療紀錄之應用:可行性研究
Tomography 2025-04-25
Creating, anonymizing and evaluating the first medical language model pre-trained on Dutch Electronic Health Records: MedRoBERTa.nl.
建立、去識別化與評估首個以荷蘭電子健康紀錄預訓練的醫學語言模型:MedRoBERTa.nl
Artif Intell Med 2025-06-05
這篇文章介紹 MedRoBERTa.nl,是首個專為荷蘭語電子健康紀錄(EHR)訓練的大型語言模型。作者用匿名化的 EHR 資料預訓練模型,讓它更懂醫療用語。文中說明模型開發、資料匿名化流程,並和其他語言模型做比較。結果顯示,MedRoBERTa.nl 能提升醫療文本分析效果,有助於治療和病患康復相關研究。
PubMedDOI