Al-Sheikh Mayss, Tepelus Tudor Cosmin, Nazikyan Tigran, Sadda SriniVas R
Doheny Image Reading Center, Doheny Eye Institute, Los Angeles, California, USA.
Department of Ophthalmology, David Geffen School of Medicine, University of California, Los Angeles, California, USA.
Br J Ophthalmol. 2017 Apr;101(4):449-452. doi: 10.1136/bjophthalmol-2016-308764. Epub 2016 Jul 22.
To determine the repeatability of optical coherence tomography angiography (OCT-A)-derived automated vessel density measurements in the superficial retinal layer (SRL) and deep retinal layer (DRL) in healthy individuals.
41 eyes of 21 healthy individuals were included in this study. OCT-A was performed using the NIDEK RS-3000 Advance with prototype OCT-A software. Each subject underwent two scanning sessions with 5-10 min in between. En face images of the retinal vasculature were generated using the default segmentation for SRL and DRL. Automated vessel density measurements were compared between the two sessions. The intraclass correlation coefficients (ICCs) and coefficient of repeatability were used as a measure for repeatability.
The mean±SD age of the subjects was 36.8±7.5 years. Overall vessel density measured in the first and second session was 19.43±3.10 and 19.72±3.78 for the SRL, and 34.67±1.801 and 34.55±1.64 for the DRL, respectively. The mean difference in vessel density between sessions was -0.3 (95% CI 3.3 to -3.9) for the SRL and 0.1 (95% CI 2.6 to -2.5) for the DRL. The two measurements were highly correlated with an ICC=0.90 (p=0.33) for the SRL and 0.83 (p=0.589) for the DRL. The coefficient of variation was 0.052 for the SRL and 0.02 for the DRL.
Commercial automated vessel density measurements using OCT-A showed excellent repeatability in healthy individuals. Although repeatability will also need to be established in the setting of disease, the level of reproducibility should be useful for assessing the significance of differences in capillary density over time or under different conditions.
确定光学相干断层扫描血管造影(OCT-A)在健康个体视网膜浅层(SRL)和深层(DRL)中自动测量血管密度的可重复性。
本研究纳入了21名健康个体的41只眼睛。使用配备原型OCT-A软件的NIDEK RS-3000 Advance进行OCT-A检查。每位受试者进行两次扫描,中间间隔5 - 10分钟。使用SRL和DRL的默认分割生成视网膜血管的正面图像。比较两次扫描之间的自动血管密度测量结果。组内相关系数(ICC)和重复性系数用作可重复性的度量。
受试者的平均年龄±标准差为36.8±7.5岁。SRL在第一次和第二次扫描中测量的总体血管密度分别为19.43±3.10和19.72±3.78,DRL分别为34.67±1.801和34.55±1.64。SRL两次扫描之间血管密度的平均差异为-0.3(95%置信区间3.3至-3.9),DRL为0.1(95%置信区间2.6至-2.5)。两次测量高度相关,SRL的ICC = 0.90(p = 0.33),DRL的ICC = 0.83(p = 0.589)。SRL的变异系数为0.052,DRL为0.02。
使用OCT-A进行的商业自动血管密度测量在健康个体中显示出极好的可重复性。尽管在疾病背景下也需要确定可重复性,但这种可重复性水平对于评估不同时间或不同条件下毛细血管密度差异的意义应该是有用的。