关键词: 冠心病；女性；模型, 统计学

Abstract: ObjectiveTo build the mathematical estimation model with multiple physiological and biochemical factors in order to improve the potency in diagnosing coronary heart disease (CHD) in female. MethodsThrough the correlation analysis and ROC curve， the factors which had significant relationship with CHD were confirmed. The mathematical estimation model was built，the diagnostic efficiency was confirmed by ROC curve and the repeatability was performed in patients. Results①There were significant differences between test group and control group except for LDLC， APOB and Lp（a）; ②There were very strong relationship between CHD and age， TG， HCY， GGT， HDLC， APOA; ③In ROC curve，the areas of five indexes such  as age，HCY， GGT， HDLC and APOA were more than 0.75; ④A total of six mathematical models were built  and the area of the sixth model was 0.955; ⑤The mathematical estimation model can be repeated to validate  the patients. ConclusionThe mathematical estimation models with multiple factors can significantly promote the potency in diagnosing CHD in female.

Key words: coronary heart disease, female, models, statistical