基于专利数据的我国医疗大语言模型应用现状与发展特征研究OA

Objective The study aims to systematically analyze published invention patent applications related to medical large language models(LLMs)in China,and to explore their application status and developmental characteristics.Methods Patentometrics and patent content analysis were employed to construct a multi-level classification framework for application domains.Quantitative analysis and visualization were conducted on disclosure trends,patent assignees,geographical distribution,keyword frequency,and application domains.Results As of April 15,2025,a total of 660 invention patent applications related to medical LLMs had been published in China.Related patents began to emerge in the second quarter of 2023 and peaked in the first quarter of 2025.Technology enterprises were the primary patent assignees,with patenting activities concentrated in Beijing,Shanghai,and Guangdong.Application scenarios were mainly clustered in two domains:medical service management and medical education and research.Key areas included clinical decision support,medical image processing,AI-aided consultation,and research assistance.Conversely,areas such as the healthcare industry,public health,and primary care remained relatively weak,indicating an imbalanced distribution of use cases.Conclusions The number of patent applications for medical LLMs in China exhibits a sharp increase,with technological innovation highly aligned with healthcare service needs.However,challenges stay in industrialization,adaptation for primary care,and system integration.It is urgent to strengthen data interoperability,expand application use cases,improve regulatory frameworks,and cultivate interdisciplinary talent to promote the sustainable development of medical LLMs.