生成式大语言模型在医疗领域的潜在典型应用与面临的挑战  

Generative Large Language Models in the Medical Domain: Potential and Typical Applications and Challenges

作　　者：颜见智 何雨鑫 骆子烨 胡晗 范士喜 汤步洲 YAN Jianzhi;HE Yuxin;LUO Ziye;HU Han;FAN Shixi;TANG Buzhou(Harbin Institute of Technology(Shenzhen),Shenzhen 518055,China;Peng Cheng Laboratory,Shenzhen 518055,China;Shenzhen Polytechnic University,Shenzhen 518055,China)

机构地区：[1]哈尔滨工业大学(深圳),深圳518055 [2]鹏城实验室,深圳518055 [3]深圳职业技术大学,深圳518055

出　　处：《医学信息学杂志》2023年第9期23-31,共9页Journal of Medical Informatics

基　　金：科技创新2030——“新一代人工智能”重大项目(项目编号:2021ZD0113402);国家自然科学基金项目(项目编号:62276082)。

摘　　要：目的/意义为快速适应新型人工智能技术发展,精准把握医疗人工智能发展方向,亟须系统地分析和梳理生成式大语言模型在医疗领域的潜在典型应用和面临的挑战。方法/过程调研分析文献与公开报道,梳理总结生成式大语言模型在医疗领域不同任务中的应用尝试和评估结果。结果/结论生成式大语言模型在医疗领域的应用逐渐增多,为医疗服务、医学研究和教育等方面提供智能辅助,同时也面临诸多挑战,如其本身存在的幻觉问题,以及数据隐私保护、伦理、结果可控性和算法可解释性等问题。Purpose/Significance In order to quickly adapt to the development of new artificial intelligence(AI)technologies and accurately grasp the development direction of medical AI,it is urgent to systematically analyze and sort out the potential and typical applications and challenges of generative large language models in the medical field.Method/Process The paper investigates and analyzes the literature and public reports,systematically summarizes the attempts and evaluation results of generative large language models in different tasks in the medical field.Result/Conclusion The application of generative large language models in the medical field is gradually increasing,covering almost all medical informatics tasks from medical information extraction,text classification,information retrieval,question and answer and dialogue to physician examination,medical record generation,medical result prediction,drug development and medical image analysis,etc.,which can provide intelligent assistance for medical services,medical research and education.At the same time,the application of generative large language models in the medical field also faces many challenges,such as the hallucination problem,data privacy protection,ethics,controllability of results and interpretability of algorithms.

关 键 词：生成式大语言模型 医疗 人工智能 

分 类 号：R-058[医药卫生]