Comparative analysis of large language models in clinical diagnosis: performance evaluation across common and complex medical cases.
大型語言模型於臨床診斷的比較分析:於常見與複雜醫療案例中的表現評估
JAMIA Open 2025-06-13
Real-time automated billing for tobacco treatment: developing and validating a scalable machine learning approach.
即時自動化菸草治療計費:開發與驗證可擴展的機器學習方法
JAMIA Open 2025-06-13
Annotation of biological samples data to standard ontologies with support from large language models.
結合大型語言模型輔助,將生物樣本數據註釋至標準本體的研究
Comput Struct Biotechnol J 2025-06-13
Exploring the medical ethical limitations of GPT-4 in clinical decision-making scenarios: a pilot survey.
GPT-4 在臨床決策情境中醫學倫理限制之探討:初步調查
Front Public Health 2025-06-13
Answering real-world clinical questions using large language model, retrieval-augmented generation, and agentic systems.
運用大型語言模型、檢索增強生成與代理系統來解答真實世界臨床問題
Digit Health 2025-06-13
Artificial intelligence in perioperative medicine education: A feasibility test of case-based learning.
人工智慧於圍手術期醫學教育中的應用:以案例導向學習進行可行性測試
J Perioper Pract 2025-06-13
Evaluating the Reliability and Quality of Sarcoidosis-Related Information Provided by AI Chatbots.
AI 聊天機器人所提供有關 Sarcoidosis 資訊之可靠性與品質評估
Healthcare (Basel) 2025-06-13
這項研究發現,採用檢索增強技術的AI聊天機器人(如ChatGPT-4o Deep Research等)在提供結節病資訊時,比一般AI更準確可靠。不過,他們的回答多半太艱深,病人不易看懂,而且給的實際建議也不夠明確。整體來說,AI雖然能提供高品質資訊,但在易讀性和實用性上還有進步空間。
相關文章PubMedDOI推理
The Usability of Neurological Occupational Therapy Case Studies Generated by ChatGPT.
由 ChatGPT 生成的神經學職能治療個案研究之可用性
Healthcare (Basel) 2025-06-13
**重點摘要:**
這項研究評估了使用 ChatGPT 來產生職能治療個案研究的可行性。五個由 AI 生成的神經學個案由十位專家進行審查,專家們對這些個案在真實性、完整性以及教育價值方面都給予高度評價。雖然專家認為這些個案很有幫助,但它們對學生學習的實際影響還需要進一步測試。
相關文章PubMedDOI推理