用数字体图像相关方法研究外源性应力作用下植物细胞骨架的力学行为

Digital volume correlation (DVC) is an effective technical approach which can quantitatively characterize the 3D dynamic displacement of cytoskeleton under mechanical stress. Digital volume correlation method based on finite element formulation avoids the misconvergence phenomenon which is caused with the error of sub-pixel interpolation. However, because of large number of the degree of freedom, FE-based DVC has low computing efficiency. Aiming at these disadvantages, the Proper Generalized Decomposition (PGD) and Multigrid method (MG) were proposed. The algorithm gives an effective solution for the question about redundant operations and low calculating speed. The multigrid PGD-DVC method provides an effective test method for study the mechanical responses of plant cytoskeleton. Combined the proposed technology with the confocal laser scanning microscope (CLSM) and atomic force microscope (AFM), explore the dynamic displacement response characteristics of plant cytoskeleton under exogenous mechanical stress, determine the stress scale and the difference that results in The results are valuable in understanding the underlying mechanism of mechanical signal transduction.

数字体图像相关(DVC)是一种能够定量表征细胞骨架在机械力应激响应的三维位移变化的方法，但是采用有限元位移模式的数字体图像相关方法由于自由度过多造成收敛速度缓慢，计算效率低。本项目将特征广义分解(PGD)方法和多重网格(MG)方法引入并发展数字体图像相关方法，能够有效地降低系统自由度和加快收敛速度，达到提高运算效率的目的，为植物细胞骨架的受力信号响应及其力学行为研究提供精确高效的测试手段。将该方法和激光共聚焦显微镜(CLSM)与原子力显微镜(AFM)联机平台相结合，探索植物细胞骨架在外源性机械应力作用下的动态位移响应，确定引起细胞骨架重排的应力规模，以及不同应力水平下细胞骨架位移场的差异性，有助于理解细胞骨架在机械力信号转导中的作用机理。

本项目将特征广义分解(PGD)方法和多重网格(MG)方法引入数字体图像相关方法，有效地降低系统自由度，并加快收敛速度，提高了运算效率。将该方法和激光共聚焦显微镜(CLSM)与原子力显微镜(AFM)相结合，探索植物细胞骨架在外源性机械应力作用下的动态位移响应，试验提供了可控微量应力加载条件下实时观测细胞骨架动态的可能性。提出了一种将熵-灰度梯度比作为指标的三维数字体散斑结构的评价方法，定量的评价分析通过激光共聚焦显微镜获得的三维数字体散斑图的质量及其所引起的误差。针对数字图像相关方法，本课题还提出自适应参数调整方法和邻域二值散斑方法。