Thepass Gijs, Lemij Hans G, Vermeer Koenraad A
*Rotterdam Ophthalmic Institute †Glaucoma Service, Rotterdam Eye Hospital, Rotterdam, The Netherlands.
J Glaucoma. 2017 Nov;26(11):1001-1009. doi: 10.1097/IJG.0000000000000764.
The purpose of this study is to explore the attenuation coefficient (AC) of the retinal nerve fiber layer (RNFL) in spectral domain optical coherence tomography (OCT) images, in healthy eyes and eyes affected by glaucoma. To assess the relation between RNLF AC, disease severity, RNFL thickness, visual field sensitivity threshold, spatial location and age.
We analyzed peripapillary circle scans of a clinical OCT device (Spectralis OCT, Heidelberg Engineering, Heidelberg, Germany) in 102 glaucoma patients and 90 healthy controls. The images were fully automatically converted into depth-resolved AC images. Next, the median AC within the RNFL was calculated based on the Spectralis segmentation. We compared the RNFL AC between healthy, mild, moderate and advanced glaucomatous eyes and assessed the correlation with patient characteristics such as age and visual field sensitivity threshold (HFA, Carl Zeiss Meditec, Dublin, CA) in a generalized estimating equations (GEE) model. Finally, we explored the ability to discriminate between glaucomatous and healthy eyes by RNFL AC.
Median RNFL AC decreased with increasing disease severity up to moderate glaucoma (P<0.001) in all 4 sectors around the optic nerve head. The largest relative decrease occurred in the nasal sector. The RNFL AC (AUC, 0.834±0.028) effectively discriminated healthy from glaucomatous eyes, although RNFL thickness (AUC, 0.975±0.013) performed even better (P<0.001). Prediction of visual field sensitivity improved significantly when RNFL thickness was augmented with RNFL AC as covariates (P<0.001).
This study demonstrated that RNFL AC provides complementary information on the RNFL's health compared with RNFL thickness measurements alone.
本研究旨在探讨健康眼睛和青光眼患者眼睛在光谱域光学相干断层扫描（OCT）图像中视网膜神经纤维层（RNFL）的衰减系数（AC）。评估视网膜神经纤维层AC与疾病严重程度、RNFL厚度、视野敏感度阈值、空间位置和年龄之间的关系。
我们分析了102例青光眼患者和90例健康对照者的临床OCT设备（德国海德堡工程公司的Spectralis OCT）的视乳头周围环形扫描图像。这些图像被全自动转换为深度分辨的AC图像。接下来，根据Spectralis分割计算RNFL内的AC中位数。我们比较了健康、轻度、中度和重度青光眼患者眼睛的RNFL AC，并在广义估计方程（GEE）模型中评估了其与年龄和视野敏感度阈值（美国加利福尼亚州都柏林卡尔蔡司医疗技术公司的HFA）等患者特征的相关性。最后，我们探讨了通过RNFL AC区分青光眼患者和健康人眼睛的能力。
在视神经乳头周围的所有4个象限中，直至中度青光眼，RNFL AC中位数随疾病严重程度增加而降低（P<0.001）。最大的相对降低发生在鼻侧象限。RNFL AC（曲线下面积，0.834±0.028）能有效区分健康人和青光眼患者的眼睛，尽管RNFL厚度（曲线下面积，0.975±0.013）表现更好（P<0.001）。当将RNFL厚度与RNFL AC作为协变量时，视野敏感度的预测有显著改善（P<0.001）。
本研究表明，与单独测量RNFL厚度相比，RNFL AC能提供关于RNFL健康状况的补充信息。
