Breidenbach K, Neppert B, Dannheim F, Hellner K A
Universitäts-Augenklinik Hamburg.
Ophthalmologe. 1996 Aug;93(4):451-5.
According to current opinion, the pattern electroretinogram (ERG) is generated in the retinal ganglion cells. Therefore, the amplitude is expected to be reduced in diseases of the optic nerve. This leads to the conclusion that the pattern ERG might be applied in detecting ganglion cell loss in the routine clinical diagnosis of chronic open-angle glaucoma. Pattern ERGs were recorded in 15 patients with defined chronic open-angle glaucoma and 21 control individuals using a DTL fiber electrode at two different check sizes. A Fourier analysis of the ERG results reduces the subjectivity when analysing the amplitudes on screen and implies the possibility of multicentric comparability. The reliability of coordinating a value of a future patient to the correct group was estimated to 82% for the Fourier data by the bivariate discriminant analysis, based on the amplitudes of both check sizes. In this way, the relatively high interindividual variability will be partly overcome. The ability of the ERG amplitudes to separate glaucoma patients from healthy persons was determined by the AUC value, using a combination of sensitivity and specificity extracted from the ROC curves. The AUC value has its maximum at 1.0. We found an AUC value of 0.8, which confirmed the information given by the pattern ERG amplitudes. Healthy individuals and glaucoma patients cannot be separated completely by the procedure but, combined with additional clinical parameters, further validation can be achieved.
根据目前的观点，图形视网膜电图（ERG）是由视网膜神经节细胞产生的。因此，在视神经疾病中其振幅预计会降低。这就得出结论，图形ERG可能应用于慢性开角型青光眼的常规临床诊断中检测神经节细胞丢失。使用DTL纤维电极在两种不同的检查尺寸下，对15例明确诊断为慢性开角型青光眼的患者和21名对照个体进行了图形ERG记录。对ERG结果进行傅里叶分析可降低在屏幕上分析振幅时的主观性，并意味着多中心可比性的可能性。基于两种检查尺寸的振幅，通过双变量判别分析估计，对于傅里叶数据，将未来患者的值正确归类到相应组别的可靠性为82%。通过这种方式，可部分克服相对较高的个体间变异性。使用从ROC曲线中提取的敏感性和特异性的组合，通过AUC值确定ERG振幅区分青光眼患者和健康人的能力。AUC值的最大值为1.0。我们发现AUC值为0.8，这证实了图形ERG振幅所提供的信息。通过该方法不能完全区分健康个体和青光眼患者，但结合其他临床参数可实现进一步验证。
