In vivo digital retinal pathology: imaging retina with molecular specific contrast and cellular resolution

我的研究计划的首要目标是开发新的成像工具和技术，用于活体视网膜成像，使活体动物模型中的疾病机制能够进行动态的细胞和分子分析。这些先进的生物显微镜工具不仅有可能促进再生医学和干细胞疗法的发展，而且还有可能产生高分辨率显微镜的新概念。 目标1.本研究项目的主要目的是开发和推进可见光一体化(VAIO)视网膜成像技术，包括可见光照明的视网膜光学相干层析成像(OCT)系统、自适应光学(AO)和偏振分集检测(PDD)。使用VAIO-OCT，单个视网膜层的形态细节将在3D各向同性细胞分辨率下以多个光学对比度可视化，从而实现非侵入性动态视网膜分子成像能力。 目标2.我的研究计划的第二个补充目标是开发新的计算图像分析和机器学习工具，这些工具是专门为数字视网膜病理定制的。通过基于像素的空间邻近度和多个光学特征相似性对像素进行聚类，可以从Vaio-OCT图像中获得视网膜组织的数字染色，这将有助于突出视网膜内部的不同组织组，并增强其对比度。这一目标将从检测、分割、特征提取和组织分类的角度实现视网膜病理图像的预测性建模。

The overarching goal of my research program is to develop novel imaging tools and technology for in vivo retinal imaging which will enable dynamic cellular and molecular analysis of disease mechanism in living animal models. These advanced bio-microscopic tools have potential to contribute not only to the development of regenerative medicine and stem cell therapies, but also to generate novel concepts for high resolution microscopy. Objective 1. The principle aim of my research program is to develop and advance Visible light All-in-One (VAiO) retinal imaging technology that comprises a retinal Optical Coherence Tomography (OCT) system with visible light illumination, Adaptive Optics (AO), and Polarization Diversity Detection (PDD). Using VAiO-OCT the morphological detail of individual retinal layers will be visualized with multiple optical contrasts at a 3D isotropic cellular resolution to enable non-invasive dynamic retinal molecular imaging capabilities. Objective 2. A complementary second objective of my research program is to develop novel computational image analysis and machine learning tools that are tailored specifically for digital retinal pathology. By clustering pixels based on their spatial proximity and multiple optical feature similarity, a digital stain' for retinal tissues from VAiO-OCT image can be achieved, which would be useful for highlighting different tissue groups inside of the retina and enhance their contrast. This objective will achieve predictive modeling of retinal pathology images from a detection, segmentation, feature extraction, and tissue classification perspective.