School of Engineering, RMIT University, Bundoora, VIC, Australia.
School of Engineering, RMIT University, Bundoora, VIC, Australia.
Comput Biol Med. 2021 Aug;135:104594. doi: 10.1016/j.compbiomed.2021.104594. Epub 2021 Jun 22.
This research work proposes a novel method for realistic and real-time modelling of deformable biological tissues by the combination of the traditional finite element method (FEM) with constrained Kalman filtering. This methodology transforms the problem of deformation modelling into a problem of constrained filtering to estimate physical tissue deformation online. It discretises the deformation of biological tissues in 3D space according to linear elasticity using FEM. On the basis of this, a constrained Kalman filter is derived to dynamically compute mechanical deformation of biological tissues by minimizing the error between estimated reaction forces and applied mechanical load. The proposed method solves the disadvantage of costly computation in FEM while inheriting the superiority of physical fidelity.
本研究工作提出了一种新的方法，通过将传统的有限元方法（FEM）与约束卡尔曼滤波相结合，对可变形生物组织进行真实和实时建模。该方法将变形建模问题转化为约束滤波问题，以便在线估计物理组织变形。它使用 FEM 根据线性弹性对 3D 空间中的生物组织变形进行离散化。在此基础上，推导出约束卡尔曼滤波器，通过最小化估计反作用力与施加机械载荷之间的误差，动态计算生物组织的机械变形。所提出的方法解决了 FEM 计算成本高的缺点，同时继承了物理逼真度的优势。
