Exploration of an Intelligent Evidence Achieve Mode of Evidence-Based Chinese Medicine：Take Systematic Review of Coronary Heart Disease Syndrome Research as an Example
10.13288/j.11-2166/r.2025.15.014
- VernacularTitle:智慧化中医药循证医学证据获取模式探索——以冠心病证候研究的系统综述为范例
- Author:
Qianzi CHE
1
;
Qingyang ZENG
2
;
Tian SONG
1
;
Lin CHEN
1
;
Jing WAN
2
;
Nannan SHI
1
Author Information
1. Institute of Basic Research in Clinical Medicine，China Academy of Chinese Medical Sciences，Beijing，100700
2. Beijing University of Chemical Technology
- Publication Type:Journal Article
- Keywords:
artificial intelligence;
coronary heart disease;
syndrome;
literature screening;
data extraction
- From:
Journal of Traditional Chinese Medicine
2025;66(15):1597-1603
- CountryChina
- Language:Chinese
-
Abstract:
ObjectiveTo construct an intelligent model for literature screening， retrieval， and data extraction with a systematic review of coronary heart disease syndromes as an example， so as to improve the efficiency of evidence-based Chinese medicine research. MethodsBased on China National Knowledge Infrastructure （CNKI）， VIP and Wanfang Data Resource System， the articles related to coronary heart disease syndrome research published from January 1， 2000 to December 31， 2023 were retrieved. Automated tools were used to batch retrieve paper metadata. Using text similarity algorithms， papers were merged， deduplicated， and subjected to preliminary screening based on titles and abstracts. Further screening was performed using object detection and image processing technologies on the full texts and statistical tables. Natural language processing （NLP） techniques and pre-trained models were applied to extract information. ResultsThe initial search retrieved 56 255 coronary heart disease syndrome-related articles. By artificial intelligence-assisted preliminary and secondary screening， the manual verification scope was narrowed to 1075 articles. Ultimately， 646 coronary heart disease syndrome related studies were included manually. With accuracy verification showing over 90% consistency in semantic recognition and element decomposition processes， we achieved data extraction and standardization processing for both basic literature information and 38 syndrome element statistics. ConclusionBy incorporating natural language processing， pre-trained models， artificial intelligence image processing and other technologies， this study enabled efficient retrieval， screening and standardized data extraction of Chinese medicine research literature.
