Bibliometric analysis on research hotspots and evolutionary trends of artificial intelligence application in traditional Chinese medicine diagnosis

Objective

To explore the development and research hotspots on the application of artificial intelligence (AI) in traditional Chinese medicine (TCM) diagnosis and predict research trends in the area.

Methods

All articles were retrieved from China National Knowledge Infrastructure (CNKI), Wanfang Data (Wanfang), China Science and Technology Journal Database (VIP), and Web of Science Core Collection (WoSCC). All related papers published in journals from the foundation of the databases to December 31, 2022 were included. NoteExpress, Co-Occurrence (COOC), VOSviewer, and CiteSpace were used to visualize data about publication volumes, journals, authors, research institutions, and keywords as well as to analyze hotspots trending topics in the field.

Results

A total of 686 articles were retrieved from the databases, among which 610 papers were published in Chinese and 76 in English. In terms of the journals in which these papers were published, 238 of them were Chinese journals and 52 were English ones. The number of the papers published in journals presented a slow growth. According to the results from Chinese article analysis, WANG Yiqin from Shanghai University of Traditional Chinese Medicine published the most papers in the field. The authors of Chinese papers belonged to six long-term research teams, led by WANG Yiqin and XU Jiatuo (Shanghai University of Traditional Chinese Medicine), WEI Yuke (Guangdong University of Technology), LI Gang (Tianjin University), XI Guangcheng (Institute of Automation of the Chinese Academy of Sciences), and NIU Xin (Beijing University of Chinese Medicine), respectively. In accordance with results from English paper analysis, four authors equally publishing the most papers were YAN Haixia, HU Xiaojuan, and JIANG Tao (Shanghai University of Traditional Chinese Medicine), and WEN Chuanbiao (Chengdu University of Traditional Chinese Medicine). The authors of English papers were from two major research teams in the field of Shanghai University of Traditional Chinese Medicine. Currently, research hotspots on AI such as neural networks, data mining, machine learning, feature recognition, image processing, and expert systems, have been centered on tongue diagnosis, pulse diagnosis, and syndrome research in TCM. Additionally, it was found that research on the topic was gradually evolving from explorations of a single diagnosis method to investigations on the combination of multiple TCM diagnosis methods.

Conclusion

Research on AI application in TCM diagnosis is still in a slowly growing stage. As technology develops, AI has been applied to many aspects of TCM diagnosis. Therefore, how to combine the two for improving TCM diagnosis is something worthy of our brainstorming and exploring.