论文部分内容阅读
对108例正常人8项肺功能指标与性别、年龄、身长、体重四个因子作逐步回归分析,进行了因子分析,并分别建立了多元回归方程,计算出多元相关系数(R),和多元相关指数(R~2),依据R、R~2大小判断与8项肺功能指标有关因子,回归分析结果显示FEF75-85%V_(25)V_(12·5·)与年龄、身长、体重有关与性别无关;FEF25-75%。V_(50)与性别、年龄有关而与身长、体重无关;FEFV_1与性别、年龄、身长有关而与体重无关。认为目前各家仅从性别、年龄、身长,体重有关因子求出预测正常值的方程,其精确性仍不够令人满意。本文利用电子计算机采用逐步回归分析对影响肺功能指标的因子进行剔选的方法,今后在类似的临床研究中有较高的应用价值。
Eight lung function indicators of 108 normal subjects were stepwisely regressed with 8 factors including gender, age, length, and weight, and factor analysis was performed. Multivariate regression equations were established to calculate multivariate correlation coefficients (R) and multivariate factors. Correlation index (R~2), which is based on R, R~2 size and 8 factors related to lung function index. Regression analysis results show that FEF75-85%V_(25)V_(12·5·) and age, length, weight There is no relationship with gender; FEF25-75%. V_(50) is related to gender and age but not to body length or body weight. FEFV_1 is related to gender, age, and body length and has nothing to do with body weight. It is believed that the equations for predicting normal values are only determined by the factors related to gender, age, length and weight, and their accuracy is still not satisfactory. In this paper, we use computer to use stepwise regression analysis to select the factors that affect the index of pulmonary function. It will have higher application value in similar clinical research.