Comb Light Source/Imaging spectrometer for advanced Spectral Domain Optical Coherence Tomography

利用光学相干断层成像技术进行功能成像的研究越来越受到人们的关注 （OCT）。这些功能成像方法中的许多方法依赖于OCT的精密相位灵敏度 干涉仪这些包括成熟的技术，如多普勒OCT，用于测量血流，以及 新兴的应用，如用于测量机械性能的OCT弹性成像和磁动OCT 和用于分子成像的光热OCT。我们自己的兴趣在于振动测量领域。 扫频激光系统是有利的，因为它们使得能够使用马赫-曾德尔型干涉仪， 平衡探测器这提供了共模噪声、DC分量和 自相关伪影。最先进的商用扫频激光器也具有非常长的 相干长度（>1 m），因此作为深度的函数的信号滚降是可忽略的。综合这些品质 导致具有较少伪影和增加的信噪比以及伴随的高相位稳定性的图像 和振动灵敏度。大多数商业上可获得的扫描激光器在激光扫描中遭受不稳定性 使得每个扫描都需要被校准以保持高相位稳定性。自己和他人 已经开发出实现这一点的方法，但是以大量的硬件复杂性为代价， 复杂的算法，以确保波数作为时间的函数，k（t），是精确地知道每个 扫根据我们的经验，系统、环境的微小变化，以及激光器的正常老化， 频繁的“调整”系统以保持最高的相位稳定性。一家公司开发了一种扫描式 激光，其中k（t）是线性的并且高度可再现。我们已经表明，该源提供高相位- 不需要复杂的硬件/软件算法。这种激光系统的缺点是 其有限的光谱带宽（通常为90-95 nm）和其有限的可用性。另一方面，在一项研究中， 基于光谱仪的系统提供高相位稳定性，但通常具有相对短的相干长度（1-3 mm），并且没有平衡检测的优点。然而，它们可以提供宽的光谱带宽 伴随着更高的分辨率，同时保持高的相位稳定性。 在这里，我们建议开发一种新的梳状光源光谱仪为基础的系统，具有所有的优点 但是具有基于光谱仪的系统的固有相位稳定性。目标1：制定 梳状激光源，中心波长约为1300 nm，带宽为125-150 nm，1024条离散线超过约200 nm.每条线的相干长度至少为8 cm，提供4 cm 3 dB滚降。目标2：开发一个 具有0.5的放大率的成像光谱仪，其在200 nm上线性地分散光（以k计 使用定制的复合棱镜，具有147 kHz的线速率。目标3：整合光源 和光谱仪组成的平衡谱域OCT系统，验证系统性能，并开发FPGA代码 用于差分信号的流水线实时计算。

There is a growing interest in the development of functional imaging with Optical Coherence Tomography (OCT). Many of these approaches to functional imaging rely on the exquisite phase sensitivity of the OCT interferometer. These include well established techniques like Doppler OCT for measuring blood flow as well as emerging applications like OCT elastography for measuring mechanical properties, and Magnetomotive OCT and Photothermal OCT for molecular imaging. Our own interest lies in the area of vibrometry. Swept laser systems are advantageous because they enable the use of Mach-Zehnder type interferometers with balanced detectors. This provides for cancelation of common mode noise, the DC component, and autocorrelation artifacts. The most advanced commercially available swept lasers also have a very long coherence length (>1 m), hence signal roll-off as a function of depth is negligible. Taken together these qualities result in images that have fewer artifacts and increased signal-to-noise with concomitant high phase stability and vibrational sensitivity. Most commercially available swept lasers suffer from instabilities in the laser sweep such that every sweep needs to be calibrated in order to maintain high phase-stability. Ourselves and others have developed methods to accomplish this, but at the expense of substantial hardware complexity and sophisticated algorithms to ensure that wavenumber as a function of time, k(t), is known precisely for each sweep. In our experience, small changes in the system, environment, and the normal aging of the laser, forces frequent “tweaking” of the system to maintain the highest phase-stability. One company has developed a swept laser, where k(t) is linear and highly reproducible. We have shown that this source provides high phase- stability without the need of the complex hardware/software algorithms. The drawback to this laser system is its limited spectral bandwidth (typically 90-95 nm) and its limited availability. On the other hand, spectrometer based systems offer high-phase stability, but typically with relatively short coherence lengths (1-3 mm) and without the advantages of balanced detection. However, they can provide wide spectral bandwidth with concomitant higher resolution while maintaining high phase-stability. Here we propose to develop a novel comb light source spectrometer based system that has all of the advantages of a swept laser system, but with the inherent phase-stability of a spectrometer based system. Aim 1: Develop a comb laser source centered at ~1300 nm, with a bandwidth of 125-150 nm, and 1024 discrete lines over ~200 nm. The coherence length of each line will be at least 8 cm, providing a 4 cm 3 dB roll-off. Aim 2: Develop an imaging spectrometer with a magnification of 0.5 that disperses the light linearly (in k) over the 200 nm bandwidth using a custom compound prism, having a line rate of 147 kHz. Aim 3: Integrate the light source and spectrometer into a balanced spectral domain OCT system, validate performance, and develop FPGA code for pipelined real-time computation of the differential signal.