Pentacam眼前节分析系统对早期圆锥角膜前后表面曲率及高度的研究

摘要

Objective To evalute the anterior and posterior corneal curvature and elevation parameters provided by Scheimpflug device , which can effectively discriminated between subclinical keratoconus, keratoconus eyes, and normal eyes and to determin which acts on the diagnosis of keratoconus. Methods Retrospective case series. Sixteen eyes of 16 patients with clinical keratoconus, 19 eyes of 19 patients with subclinical keratoconus, and 56 eyes of 56 normal subjects. One eye from each patient was examined, the anterior and posterior corneal curvature and elevation parameters are provided by Scheimpflug device. The Mann-Whitney test, receiver-operating-characteristic ( ROC ) curves, and Partial Least Squares ( PLS )were used to compare the mean measurements and to evaluate the sensitivity and specificity of the parameters or constructed models. Results All of the parameters which are composed of anterior and posterior corneal curvature and elevation have more significant changes in eyes with subclinical or clinical keratoconus than in normal eyes ( P <0.05 ). The ROC curve analysis showed high overall predictive accuracy of all of the parameters in discriminating ectatic corneas from normal corneas. Partial Least Squares( PLS ) analysis showed that the goodness of fit of a model using a combination of 11 parameters of the corneal anterior surface was best in discriminating keratoconus or subclinical keratoconus eyes from normal eyes. Conclusion Posterior and anterior elevation, and curvature parameters measured by the Pentacam camera can effectively discriminate keratoconus from normal corneas serving as a useful diagnostic tool for disease staging.%目的 应用Pentacam眼前节分析系统测得的角膜前后表面曲率、高度参数来探讨正常角膜与临床期圆锥角膜和亚临床期圆锥角膜的关系,以及Pentacam眼前节分析系统在圆锥角膜诊断中的作用.方法 回顾性病例系列.选取临床期圆锥角膜患者16例(16只眼),亚临床期圆锥角膜19例(19只眼),对照组56例(56只眼)正常人群.每例患者选取一只眼,将Pentacam眼前节分析系统测得角膜前后表面曲率和前后表面高度参数进行差异检验以及受试者工作特征(ROC)曲线分析,初步筛选出有用参数,进一步进行偏最小二乘(PLS)分析,确定出最终诊断模型,并利用ROC曲线方法对模型作出评价.结果 相对于正常患者,临床期患者与亚临床期患者的角膜前后表面曲率、高度参数有明显的变化.ROC曲线分析显示所有参数对于将圆锥角膜从正常眼中区分出来有较高的预测精准度.应用PLS进行变量筛选并将VIP大于1.0的11个参数建立模型可以有效的将临床期圆锥角膜眼或亚临床期圆锥角膜眼从正常眼中区分出来.结论 由Pentacam眼前节分析系统测得的角膜前后表面的曲率、高度参数可以有效地将圆锥角膜从正常角膜中区分出来,并且能作为疾病分期的诊断工具.