摘要
目的:分析冠心病痰瘀互结证脉诊信息特征,为冠心病证型辨识提供客观依据。方法:采集的冠心病患者瘀互结证组和非痰瘀互结证组(对照组)脉诊信息进行分析。采用无监督学习的主成分分析和有监督学习的LS回归、Lasso回归分析方法进行分类判决。结果:第1、9主成分所涉及的右手寸脉时域参数;双手频域相位参数降维拟合具有统计学t检验和f检验差异,有监督得到LS模型冠心病痰瘀互结组与非痰瘀互结证整体组判别准确率为74%;冠心病痰瘀互结组与痰阻热蕴、气滞血瘀证、气虚血瘀证、气阴两虚证、阳虚寒凝证各证型各组LS识别模型判别准确率为69%~78%,Lasso回归分析下,冠心病痰瘀互结证与非痰瘀互结证2组之间模型判别准确率分别为74%。构成2种证型脉诊信息差异的主要参数集中在左寸相位3、左寸相位6、左关相位6、右关相位6、右尺相位12等频域相位参数;冠心病痰瘀互结证与非痰瘀互结证2组之间模型判别在右寸S2时域参数特征明显。结论:脉诊数据与中医理论对应分析后,冠心病痰瘀互结证与非痰瘀互结证的证型脉象频域信息参数区别主要差异体现在双手寸脉、尺脉。尤其是在冠心病痰瘀互结证痰浊、瘀特征辨识时,以右寸S2时域差异尤为明显。反映出“肺主治节”在气血精液输布的重要作用。
Objective:To analyze the pulse diagnosis information characteristics of coronary heart disease(CHD)with the syndrome of phlegm and blood stasis,and provide an objective basis for identifying CHD syndromes.Methods:Pulse diagnosis data from CHD patients with phlegm and blood stasis syndrome and without this syndrome(control group)were collected and analyzed.Unsupervised principal component analysis(PCA)and supervised learning methods including Least Squares(LS)regression and Lasso regression were used for classification and decision-making.Results:The right-hand Cun pulse time-domain parameters in the 1st and 9th principal components,along with the frequency-domain phase parameters of both hands,showed statistically significant differences in t-tests and F-tests.The supervised LS model achieved an overall classification accuracy of 74%in distinguishing between the phlegm and blood stasis syndrome group and the control group.In distinguishing the phlegm and blood stasis syndrome group from other syndromes,including phlegm obstruction and heat accumulation,qi stagnation and blood stasis,qi deficiency and blood stasis,qi and yin deficiency,and yang deficiency and cold congealing,the LS model achieved accuracy rates of 69%to 78%.In Lasso regression analysis,the classification accuracy between the phlegm and blood stasis syndrome group and the control group was 74%.The main parameters distinguishing the two syndromes were concentrated in frequency-domain phase parameters,including left Cun phase 3,left Cun phase 6,left Guan phase 6,right Guan phase 6,and right Chi phase 12.The most significant parameter distinguishing the phlegm and blood stasis syndrome group from the control group was the right Cun S2 time-domain feature.Conclusion:After corresponding analysis of pulse diagnosis data and traditional Chinese medicine(TCM)theory,the main differences in pulse information between phlegm and blood stasis syndrome and non-phlegm and blood stasis syndrome were found in the frequency-domain parameters of both the Cun and Chi pulses.Particularly,the right Cun S2 time-domain difference was significant in identifying the characteristics of phlegm and blood stasis in CHD with phlegm turbidity and blood stasis syndrome.This reflects the important role of the“lung governing management and regulation”in the distribution of qi,blood,and essence.
作者
李阳
刘佳
莫国凤
刘跞
王星雅
常娜娜
汪南玥
LI Yang;LIU Jia;MO Guofeng;LIU Luo;WANG Xingya;CHANG Nana;WANG Nanyue(Basic Medical College of Shaanxi University of Chinese Medicine,Xianyang 712046,China;Medical Experiment Center of China Academy of Chinese Medical Sciences,Beijing 100700,China;Basic Research on Prevention and Treatment of Major Diseases of Traditional Chinese Medicine Beijing Key Laboratory,Beijing 100700,China;Jiangxi Institute of Traditional Chinese Medicine and Health Industry,China Academy of Chinese Medical Sciences,Nanchang 330000,China)
出处
《世界中医药》
北大核心
2024年第19期3008-3014,共7页
World Chinese Medicine
基金
国家重点研发计划项目(2018YFC1707605)
中国中医科学院科技创新工程项目(CI2021A05207)
北京市科技新星交叉合作课题(BJJC2022001)
中央级公益性科研院所基本科研业务费专项资金项目(JBGS2021008)。
关键词
脉象特征
主成分分析
回归分析
时域差异
监督学习
冠心病
痰瘀互结证
痰浊
肺主治节
Pulse characteristics
Principal component analysis
Regression analysis
Time-domain difference
Supervised learning
Coronary heart disease
Phlegm and blood stasis syndrome
Phlegm turbidity
Lung governing management and regulation
