Advanced optical techniques for tissue identification and visualisation

神经科学的目标是了解神经系统如何在分子、细胞和网络水平上自我组织，将分子和电信号处理成信息和行动。然而，由于我们在适当的水平(分子、细胞或系统)观察完整神经系统的结构和操纵其功能的能力有限，这一努力受到了极大的阻碍。我们与神经生物学家密切合作，创造工具，利用基于光子学的技术解决这些未得到满足的需求。这是一个漫长的过程，其根源在于展示了如何使用明智地选择的物理现象(如相干拉曼成像或照明的偏振控制)来成像以前无法获得的信息，或者如何将光束整形和光学设计中的几个概念集成在一起，可以产生具有非常理想的特性的新型设备来研究活的神经系统。NSERC发现基金的目的是支持概念验证计划，使来自物理科学和工程的学生能够开发这样的工具。这项研究计划的总体愿景是基于组织中的内源性对比机制开发新的分子成像方法，并为超大型3D成像数据集开发必要的分析工具。为此，我们将针对两个具体的问题(脂膜顺序和3D分析)，并通过需要深入了解光-物质相互作用和光学设计的项目来解决它们。首先，我们建议使用一种基于偏振敏感相干拉曼成像的成像策略来成像膜内脂类的分子顺序。简而言之，相干拉曼过程多年来一直被我们和其他人用来对髓磷脂进行成像，并且还可以产生神经系统中白质的图像。在这里，我们将利用相干拉曼过程的偏振敏感性，通过一组不同偏振下的图像来提取髓鞘中脂类的分子顺序。其次，我们将开发智能压缩策略，将大型数据集矢量化，使其更易于操作，而不会丢失信息。这将在两个测试系统中进行验证，以量化连接地图和地区分布。在我们的计划结束时，我们将提供创新成像设备和图像分析的概念和验证的证明，并将部署给其他研究人员。

The goal of neuroscience is to understand how the nervous system organizes itself at molecular, cellular and network levels to process molecular and electrical signals into information and actions. However, this endeavour is greatly impeded by our limited ability to observe the structure and manipulate the function of the intact nervous system at the appropriate level (molecular, cellular or systemic). We work in close collaboration with neurobiologists to create tools that address these unmet needs with photonics-based technology. This is a lengthy process that has its origin in demonstrating how judiciously chosen physical phenomena (such as coherent Raman imaging or polarization control of the illumination) can be used to image previously unobtainable information, or how the integration of several notions from beam shaping and optical design can lead to novel devices with highly desirable properties to study the live nervous system. The purpose of this NSERC Discovery grant is to feed the proof-of-concept program that makes possible the development of such tools with students from physical sciences and engineering. The overall vision of this research program is to develop novel molecular imaging methods based on endogenous contrast mechanisms in tissue and develop the necessary analysis tools for very large 3D imaging datasets. To do so, we will target two specific problems (lipid membrane order and 3D analysis) and address them with projects that require an intimate knowledge of light-matter interactions and optical design. First, we propose to use an imaging strategy based on polarization-sensitive coherent Raman imaging to image the molecular order of lipids within membranes. Briefly, the coherent Raman process has been used for years to image the lipids of myelin, by us and others, and can yield images of the white matter in the nervous system. Here we will turn to our advantage the polarization-sensitivity of the coherent Raman process to extract the molecular order of lipids within myelin with a set of images at different polarizations. Second, we will develop intelligent compression strategies that will vectorize large datasets to make them easier to manipulate without losing information. This will be validated in two tests systems to quantify connectivity maps and regional distributions. At the end of our program, we will have delivered proofs of concept and validation of innovative imaging devices and image analysis that will be deployed to other researchers.