摘要
目的建立冠心病不稳定型心绞痛患者临床常规检测指标对阳虚证的识别模式。方法采集411例冠心病不稳定型心绞痛患者的基本资料、中医四诊信息及临床常规检测指标,对患者进行中医证候/证候要素及辨证分型的判断。经数据归一化处理后,以90个临床常规检测指标为自变量,以是否'阳虚证'作为因变量,采用卡方自动交互检测(CHAID)决策树方法自动提取阳虚证的识别规律,计算模型的敏感度、特异度和准确率。对其中212例患者进行阳虚证识别模式的外验证。结果 D-二聚体、血小板分布宽度(PDW)、血清游离四碘甲状腺原氨酸(FT4)、脂蛋白(a)[LP (a)]、氯离子(Cl^-)和凝血酶原时间(PT)共6个属性指标经筛选进入决策树模型。该模型对411例患者的测试结果显示:敏感度为74. 7%,特异度为72. 6%,准确率为73. 0%。外验证模型缺失Cl^-,模型识别准确率为81. 1%,敏感度为74. 3%,特异度为82. 5%。结论 CHAID决策树模型可以基于临床常规检测指标数据清晰、直观地进行冠心病不稳定型心绞痛阳虚证的识别,自动归纳识别规律。
Objective To build a recognition mode of Yang deficiency syndrome according to indexes of clinical routine examinations of patients with coronary heart disease unstable angina(UA).Methods The basic data,traditional Chinese medicine(TCM)four-diagnosis information and clinical routine examination indexes of 411 UA patients were collected.TCM syndrome or syndrome elements and syndrome differentiation were judged by doctors.After data normalization,90 clinical routine examination indexes were taken as independent variables,and'yang deficiency syndrome'was taken as dependent variable.The recognition rules of yang deficiency syndrome were automatically extracted by chi-square automatic interaction detection(CHAID)decision tree method,and the sensitivity,specificity and accuracy rate of the model were calculated.The verification of recognition mode of yang deficiency syndrome on212 included patients was conducted.Results D-dimer,platelet distribution width(PDW),free tetraiodothy roine T4(FT4),lipoprotein(a)(LP(a)),chloridion(CI)and prothrombin time(PT)were selected into the decision tree model.The test results of 411 patients of the model showed:sensitivity was 74.7%,specificity was 72.6%,and the accuracy rate was 73.0%.The verification of association patterns lacks Cl,and its accuracy rate was 81.1%,sensitivity was 74.3%,and specificity was 82.5%.Conclusion CHAID decision model could be used to conduct recognition of yang deficiency syndrome of coronary heart disease UA clearly and directly based on clinical routine examination indexes,and to extract recognition rules automatically.
作者
史琦
孔艳华
赵慧辉
陈建新
李友林
王伟
SHI Qi;KONG Yanhua;ZHAO Huihui;CHEN Jianxin;LI Youlin;WANG Wei(China-Japan Friendship Hospital/National Clinical Research Center for Respiratory Diseases,Beijing,100029;Beijing University of Chinese Medicine)
出处
《中医杂志》
CSCD
北大核心
2019年第15期1312-1317,共6页
Journal of Traditional Chinese Medicine
基金
国家科技重大新药创制专项(2009ZX09502-018)
国家中医药管理局重点研究室
中医药防治过敏性疾病北京市重点实验室(BZ0321)
关键词
冠心病
不稳定型心绞痛
阳虚证
卡方自动交互检测
决策树模型
coronary heart disease
unstable angina
yang deficiency syndrome
chi-square automatic interaction detection
decision tree
