基于光片照明的三维光学血管造影方法及量化分析研究

光片照明显微成像技术是近二十年发展起来的最具代表性的高质量生物成像技术之一。本项目针对现有光片照明显微成像技术需通过荧光标记进行成像这一问题，提出了基于光片照明的光学血管造影新方法，以光学造影替代荧光造影，实现了无标记的三维血管网络成像。具体研究内容包括：1）建立基于双侧光片照明的连续变焦图像采集系统，实现原始数据采集。2）建立基于时空分析的随机森林模型和U-Net网络模型，解决光片照明下非荧光标记的短时血流弱信号提取的难题；提出迭代反投影算法，建立二维血流投影图像与三维血管网络的映射关系，实现三维血管网络重建；提出血管量化参数提取算法，获得定量的血管形态参数和统计参数。3）结合仿体和在体实验，验证基于光片照明的光学血管造影及量化分析方法的有效性，并探索其适用范围。同时，研究血管形态和量化参数与肿瘤等疾病发展的关联性。本项目对发展前沿的无标记光片照明显微成像技术做出了积极地探索。

In recent 20 years, microscopic imaging with light-sheet illumination is one of the most representative high-quality bio-imaging technology. In this project, to solve the problem that the existing light-sheet microscopic imaging needs fluorescent markers, a new optical angiography method based on light-sheet illumination is proposed, in which fluorescent markers is instead of its optical characteristics and the label-free 3D vascular imaging is realized. This research contents include the fllowing sections. 1) a continuous zoom imaging acquisition system based on two-sided light-sheet illumination will be established to obtain raw data. 2) The random forest model and U-Net network model based on spatiotemporal analysis will be established to solve the problem of extracting the short-term weak blood flow signals without fluorescence markers under the illumination of light sheet; the iterative back projection algorithm will be proposed to establish the relationship between the two-dimensional blood flow projection image and the three-dimensional blood vessel network, so as to realize the three-dimensional reconstruction of the blood vessel network; the algorithm of extracting the quantitative parameters of blood vessels will be proposed to obtain the quantitative parameters of blood vessel morphology and statistical parameters. 3) innovatively proposing digital projection tomography, establishing relationship between 2D blood flow projection image and three-dimensional vascular network space; 3) Combined with the phantom and in vivo experiments, the validity of optical angiography based on light-sheet illumination and quantitative analysis will be verified, and its application scope will be explored. At the same time, the relationship between vascular morphology and quantitative parameter, and the development of tumor and other diseases will be studied. This project has made a positive exploration on the development of label-free microscopic imaging with light-sheet illumination.