基于深度学习的中医诊断方法知识图谱分析

Knowledge Atlas Analysis of TCM Diagnosis Methods Based on Deep Learning

摘要: 目的：利用CiteSpace 6.1.R6软件对基于深度学习的中医诊断方法知识图谱进行可视化分析，探索该领域的研究现状、热点及趋势。方法：以中国知网数据库为资料来源，检索2009年1月1日至2022年12月31日的相关文献，采用Excel统计分析发文量，通过CiteSpace软件进行可视化分析。结果：共获得283篇有效文献，研究经历了萌芽、发展、爆发、平台四个阶段，总体呈上升趋势，近几年关注度达高峰。共有181位作者，高发文量作者包括温川飙等五人。研究机构主要为中医药大学、医学研究所和科技理工类大学，但联系较少。关键词共现图形成8个聚类块，深度学习与传统中医诊断技术结合逐渐实现智能化，研究热点聚焦在舌诊方面。结论：深度学习算法正逐渐深入中医诊断方法中，但我国对该领域的研究和智能诊疗应用有待发展，希望未来加强团队合作，深入研究。

Abstract: Objective: Using CiteSpace 6.1.R6 software for visual analysis of the knowledge map of deep learning-based Chinese medicine diagnosis methods to explore the research status, hotspots, and trends in this field. Methods: Drawing on data from the China National Knowledge Internet (CNKI) database, relevant literature published between January 1, 2009, and December 31, 2022, was retrieved. Excel was used for statistical analysis of publication volumes, and CiteSpace software was employed for visual analysis. Results: A total of 283 valid articles were obtained, showing a general upward trend with research progressing through four stages: emergence, development, explosion, and plateau. The field has garnered significant attention in recent years, reaching a peak. There were 181 authors, with the top five authors, including Wen Chuanbiao, contributing the highest number of publications. The main research institutions included universities of traditional Chinese medicine, medical research institutes, and science and technology-oriented universities, though with limited collaboration among them. The keyword co-occurrence map formed eight clusters, indicating the gradual integration of deep learning with traditional Chinese medicine diagnosis techniques toward intelligence, with research hotspots centering on tongue diagnosis. Conclusion: Deep learning algorithms are increasingly being incorporated into Chinese medicine diagnosis methods. However, research and the application of intelligent diagnosis and treatment in this field in China still require further development. It is hoped that future efforts will enhance teamwork and deepen research.