论文部分内容阅读
本文采用逐步判别分析的多因素分析法对６２例无精子症病人的睾丸活检、睾丸大小、血清生殖激素水平（ＦＳＨ、ＬＨ、Ｔ及Ｔ／ＬＨ）等资料进行计算机统计处理，以建立无精子症临床分类和病理类型间鉴别诊断的函数模型，结果得出了临床两类间和病理五类间的判别函数式。对方程作判别显著性检验证明方程成立（Ｐ＜０．０１），利用原观察资料分别回代判别函数式，判别总符合率分别为９５．１６％、７７．４２％，说明判别效果较可靠。因此认为这些数学模型有一定的临床应用前景。
In this paper, stepwise discriminant analysis of multivariate analysis of 62 cases of azoospermic testis biopsy, testicular size, serum reproductive hormone levels (FSH, LH, T and T / LH) and other data were computerized to establish azoospermia Disease clinical classification and pathological types of differential diagnosis between the functional model, the results obtained two clinical and pathological classification between the discriminant function. The equation was tested by significant test (P <0.01). Using the original observation data, respectively, the discriminant function was substituted and the total coincidence rates were 95.16% and 77.42%, respectively, indicating that the discriminant effect was more reliable . So that these mathematical models have some clinical application prospects.