中国中西医结合杂志2025，Vol.45Issue(9)：1043-1049,7.DOI:10.7661/j.cjim.20250726.270
基于大数据与深度学习构建类风湿关节炎中医证候智能测评量表
Construction of Intelligent Evaluation Scale for Chinese Medicine Syndromes of Rheumatoid Arthritis Based on Big Data and Deep Learning
摘要
Abstract
Objective To construct an intelligent assessment scale for Chinese Medicine(CM)syndromes of rheumatoid arthritis(RA)based on data mining and deep learning technologies,and to enhance the accuracy and objectivity of CM syndrome evaluation for RA.Methods A multi-center cohort study design was adopted to collect clinical information of RA patients from 16 national CM/integrated CM and Western medicine hospitals.After data cleaning,Cronbach's alpha coefficient method,factor analysis method,and discrete trend method were used for items screening.A subjective and objective comprehensive weighting method was used to determine the weight coefficient,in which the subjective weight was determined by the expert consensus method,and the objective weight was determined by multivariate regression analysis and deep learning algorithms.Random forest algorithm was used for feature importance analysis,and random over-sampling and re-sampling algorithm over-sampling and synthetic minority over-sampling technique sample equalization processing were performed.Finally,the internal consistency reliability of the scale was evaluated by Cronbach's alpha coefficient,and the structural validity of the scale was tested by factor analysis.Results A total of 5 880 RA patients were finally included,and 16 main symptom items,30 secondary symptom items,and 30 tongue and pulse items of RA CM syndrome efficacy evaluation scale were determined.The final scale reliability coefficient was 0.788,and the structural validity coefficient was 0.735.Conclusion This scale had high reliability and validity,and each item had higher internal consistency and characteristic reflection ability,which objectively reflected the CM syndrome characteristics of RA patients.关键词
类风湿关节炎/中医证候/量表/大数据/深度学习/疗效评价Key words
rheumatoid arthritis/Chinese medical syndrome/scale/big data/deep learning/evaluation of therapeutic effect引用本文复制引用
玉冰冰,韩曼,姜泉,杨晶东,唐晓颇,张俊杰,宋梦歌,彭秋伟,赵雪琦..基于大数据与深度学习构建类风湿关节炎中医证候智能测评量表[J].中国中西医结合杂志,2025,45(9):1043-1049,7.基金项目
中央高水平中医医院临床研究和成果转化能力提升项目(No.HLCMHPP2023087) （No.HLCMHPP2023087）
中国中医科学院科技创新工程项目(No.CI2021A01503) （No.CI2021A01503）
中华中医药学会雏鹰计划中医临床青年人才研修项目(第一期)(No.CYJH2024015) （第一期）
