Mouse Modeling Techniques for Bioluminescence Tomography

描述（由申请人提供）：生物发光断层扫描（BLT）和荧光分子断层扫描（FMT）是研究小动物分子和细胞活性的两种新方法。这些成像工具有助于诊断疾病，监测治疗和促进药物设计。小鼠的几何和光学建模是BLT/FMT的一个至关重要的先决条件。该项目的目标是开发具有独特功能的新型小鼠建模技术，包括将小鼠解剖结构分解为不同“器官区域”的方案，用于基于知识的分割的可变形小鼠图谱，用于确定器官区域的光学参数以最佳匹配来自透照成像实验的测量数据的分段常数拟合程序，以及补偿由扩散近似、几何误差和其它因素引入的误差的固有能力。所有这些优点有助于提高BLT小鼠建模的准确性、效率和鲁棒性。我们的主要假设是，BLT重建质量将显着提高使用所提出的建模技术相对于目前的方法，其中光学参数取自文献。具体目标是：（1）建立活体小鼠透照成像实验原型系统;（2）建立与真实的小鼠在光子传播方面尽可能等效的小鼠数值模型，并开发用户友好的BLT应用软件包;（3）通过数值模拟和体模实验对所提出的建模技术进行了评价和验证，并在小鼠研究中展示了其优点。在这种方法中，简化的小鼠几何模型和扩散近似模型相结合，估计代表性的光学参数从测量的透照成像数据。小鼠模型将用于BLT图像重建。在这个项目完成后，提出的小鼠建模技术将得到验证，并在BLT图像重建的相关收益将被量化的源质心和总能量。

DESCRIPTION (provided by applicant): Bioluminescence tomography (BLT) and fluorescence molecular tomography (FMT) are two new modalities to study molecular and cellular activities in small animals. These imaging tools help diagnose diseases, monitor therapies and facilitate drug designs. Geometrical and optical modeling of mice is a critically important prerequisite for BLT/FMT. The goal of this project is to develop novel mouse modeling techniques with unique features including a scheme that decomposes the mouse anatomy into different "organ regions", a deformable mouse atlas for knowledge-based segmentation, a piecewise constant fitting procedure for determination of optical parameters of organ regions to optimally match measured data from transillumination imaging experiments, and an inherent capability of compensating for the errors introduced by the diffusion approximation, geometrical errors and other factors. All these advantages contribute to improve the accuracy, efficiency and robustness of the mouse modeling for BLT. Our primary hypothesis is that the BLT reconstruction quality will be significantly improved using the proposed modeling techniques relative to the current method in which optical parameters are taken from the literature. The specific aims are to (1) build a prototype system for transillumination imaging experiments of a living mouse; (2) construct the mouse numerical models, which are equivalent to the real mice as closely as possible in terms of photon propagation, and develop a user-friendly software package for BLT applications; and (3) evaluate and validate the proposed modeling techniques in numerical simulation and phantom experiments, and demonstrate their merits in mouse studies. In this approach, the simplified mouse geometrical models and diffusion approximation models are combined to estimate representative optical parameters from measured transillumination imaging data. The mouse models will be used for BLT image reconstruction. Upon the completion of this project, the proposed mouse modeling techniques will have been validated, and the associated gains in BLT image reconstruction will have been quantified in terms of source centroid and total energy.