Advanced Laser Source for High-speed Adaptive Optics OCT

描述(由申请人提供)： 物理科学公司(PSI)建议开发一种新的自适应光学成像方法，以利用增强型傅立叶域模式锁定(FDML)扫描源技术进行高速3D-视网膜映射。PSI目前正在致力于高性能、高分辨率多模式自适应光学(AO)视网膜成像仪的开发。这些系统解析和量化黄斑上的视锥感光细胞，成像和描绘视网膜各层，显示和绘制微血管和脉络膜深处的图像，这些多模式AO系统集成了扫描激光眼底检查(AOSLO)、广场成像、视网膜跟踪，据我们所知，是第一个采用1-5M扫描源的傅立叶域光学相干断层扫描(AO-SSOCT)的多模式系统，适合在广泛的临床人群中用于研究结构和功能细节。然而，在活的、凝视的眼睛中，即使是最小的运动也会相对于典型的AO扫描的小尺寸(~1到3度)被放大。目前，使用现有的谱域或扫描源技术，即使是接近300/S的高分辨率AO-OCT B扫描速率(~10倍视频速率500A线/扫描)仍然太慢，无法可靠地捕捉特定层的局部3D结构，其横向运动可以在几毫秒的线间间隔内多达几个锥体直径。这些扫描必须达到足够的速度，以充分地将AO-OCT锥体图像与AOSLO锥体马赛克相互关联，以获得真正的多模式优势。麻省理工学院研究人员最近开发的傅里叶域锁模扫描源激光技术使OCT A线速率达到370kline/S，甚至更高。因此，在高分辨率下可获得的B超速率接近阈值(&gt；1000/S)，在该阈值下，来自人眼的运动伪影不会不可挽回地破坏局部锥体图案，这是在临床有用的视网膜区域和体积上进行图像校正、对齐和马赛克生成的最佳基础。为了实现这些能力，需要在实用的多模式声纳系统和FDML或等效的扫频源技术方面取得进展。PSI及其子公司Q-Peak处于有利地位，可以同时在两条战线上行动。此外，PSI积极参与定制图形处理单元(GPU)算法和系统的开发，用于可应用于AO-OCT的实时、高速图像立方体处理。 公共卫生相关性： 高速、高分辨率眼科成像的出现，为了解和治疗眼部疾病和视力丧失开辟了新的研究途径。所提出的采用高速扫描源激光技术的高性能声学分析仪将使更多的临床医生和科学家使用强大的自适应光学技术。这些研究人员将反过来发现这种新工具在临床应用中检测疾病和监测治疗的新的和重要的用途。

DESCRIPTION (provided by applicant): Physical Sciences Inc. (PSI) proposes to develop a new adaptive optics imaging approach to high- speed 3D-retinal mapping with enhanced Fourier Domain Mode-locked (FDML) swept-source technology. PSI is currently engaged in ongoing development of high performance, high resolution multimodal adaptive optics (AO) retinal imagers. Such systems resolve and quantify cone photoreceptors across the macula, image and delineate retinal layers, reveal and map microvasculature, and image deep in the choroid, These multimodal AO systems integrate scanning laser ophthalmoscopy (AOSLO) , wide-field field imaging, retinal tracking, and, to our knowledge, are the first multimodal systems to incoprporate 1-5m swept source-based Fourier domain optical coherence tomography (AO-SSOCT) suitable for use in a broad clinical population for investigating structural and functional detail. However, in the living, fixating eye, even the smallest motions are magnified relative to the small dimensions of typical AO scan (~1 to 3 deg). At present, with existing spectral domain or swept-source technology, even high-resolution AO-OCT B-scan rates approaching 300/s (~ten times video rate @ >500 A-lines per scan) remain too slow to reliably capture the local 3D structure of particular layers, whose lateral motions can be as much as few cone diameters in the inter-line interval of several milliseconds. These scans must reach a speed sufficient to adequately cross-correlate AO-OCT cone images with AOSLO cone mosaics to derive the true multimodal advantage. The recent development of Fourier domain mode-locked [FDML] swept-source laser technology by researchers at MIT has enabled the OCT A-line rate to reach 370kline/s, and beyond. Thus, the resulting accessible B-scan rates at high resolution approach a threshold (> 1000/s) where motion artifacts from the human eye do not irretrievably disrupt the local cone pattern, which is the best basis for image correction, alignment and mosaic generation over clinically useful retinal areas and volumes. To achieve these capabilities, progress on both practical multimodal AO systems and FDML or equivalent swept-source technologies is needed. PSI and its subsidiary Q-Peak are well positioned to move on both fronts together. In addition, PSI is actively engaged in the development of customized Graphical Processing Unit (GPU) algorithms and systems for real time, high-speed image cube processing that can be applied to AO-OCT. PUBLIC HEALTH RELEVANCE: The advent of high speed, high resolution ocular imaging has opened new avenues of research into the understanding and treatment of eye disease and vision loss. The proposed high performance AO instrument with high speed swept source laser technology will allow powerful new adaptive optics technology to be used by greater numbers of clinicians and scientists. These researchers will, in turn, find novel and important uses for this new tool in clinical application for detecting disease and monitoring treatment.