Cense Barry, Koperda Erik, Brown Jeffrey M, Kocaoglu Omer P, Gao Weihua, Jonnal Ravi S, Miller Donald T
Indiana University, School of Optometry, 800 E. Atwater Avenue, Bloomington, Indiana 47405, USA .
Opt Express. 2009 Mar 2;17(5):4095-111. doi: 10.1364/oe.17.004095.
Ultrabroadband sources, such as multiplexed superluminescent diodes (SLDs) and femtosecond lasers, have been successfully employed in adaptive optics optical coherence tomography (AO-OCT) systems for ultrahigh resolution retinal imaging. The large cost differential of these sources, however, motivates the need for a performance comparison. Here, we compare the performance of a Femtolasers Integral Ti:Sapphire laser and a Superlum BroadLighter T840, using the same AO-OCT system and the same subject. In addition, we investigate the capability of our instrument equipped with the Integral to capture volume images of the fovea and adjacent regions on a second subject using the AO to control focus in the retina and custom and freeware image registration software to reduce eye motion artifacts. Monochromatic ocular aberrations were corrected with a woofer-tweeter AO system. Coherence lengths of the Integral and BroadLighter were measured in vivo at 3.2 microm and 3.3 microm, respectively. The difference in dynamic range was 5 dB, close to the expected variability of the experiment. Individual cone photoreceptors, retinal capillaries and nerve fiber bundles were distinguished in all three dimensions with both sources. The acquired retinal volumes are provided for viewing in OSA ISP, allowing the reader to data mine at the microscope level.
超宽带光源,如复用超发光二极管(SLD)和飞秒激光器,已成功应用于自适应光学光学相干断层扫描(AO-OCT)系统中,用于超高分辨率视网膜成像。然而,这些光源的成本差异巨大,这就需要对其性能进行比较。在此,我们使用同一AO-OCT系统和同一受试者,比较了Femtolasers Integral钛宝石激光器和Superlum BroadLighter T840的性能。此外,我们还研究了配备Integral的仪器使用AO控制视网膜聚焦,并利用定制和免费软件图像配准软件减少眼动伪影,以获取第二个受试者的中央凹及相邻区域体图像的能力。使用低音-高音AO系统校正单色眼像差。Integral和BroadLighter的相干长度在体内分别测得为3.2微米和3.3微米。动态范围的差异为5分贝,接近实验预期的变化范围。两种光源在所有三个维度上均能区分单个视锥光感受器、视网膜毛细血管和神经纤维束。所获取的视网膜体积数据可在OSA ISP中查看,读者能够在显微镜层面进行数据挖掘。