论文部分内容阅读
本文应用多因素统计学方法,对糖尿病患者的暗适应功能与眼部改变及全身因素间的关系进行了计算机分析及预测。从众多的因素中筛选出对暗适应功能有显著综合作用的六个因素(控制Ⅱ级的病程时间、合并肾病、视力、晶体改变、玻璃体改变、糖尿病性视网膜病变),其中后三项对暗适应功能的改变起有非常显著性影响。据此建立起预测方程。预测方程验证结果:训练样本的预测正确率为80.77%,双盲试验对检验样本的预测正确率为77.27%,表明这是一种简便、有效的助诊方法。本文对多指标逐步回归法的优点及暗适应机理进行了讨论。
In this paper, using multivariate statistical methods, diabetic patients with dark adaptation function and eye changes and systemic factors between the computer analysis and prediction. From the many factors in screening dark adaptation function has a significant role in the synthesis of six factors (control of Ⅱ duration of the disease, with nephropathy, visual acuity, changes in the crystal, vitreous changes, diabetic retinopathy), of which the latter three dark Adapt to functional changes have a very significant impact. Based on this, a prediction equation is established. Prediction equation validation results: the correct rate of the prediction of training samples was 80.77%, double-blind test of the test sample prediction accuracy of 77.27%, indicating that this is a simple and effective method of diagnosis. This article discusses the advantages of multi-index stepwise regression and the mechanism of dark adaptation.