AI 中文总结
本研究提出一种多层纤维增强连续体计算框架，结合主动应变法与增广拉格朗日接触算法，通过GetFEM有限元代码建模，首次实现三维小肠段生理蠕动与褶皱的准确模拟，为胃肠运动研究提供新范式。
AI 中文摘要
本文提出了一种新的计算框架，将肠壁建模为多层纤维增强连续体。该框架首次再现了生理运动能力，并克服了经典软组织超弹性公式中存在的大位移限制（自接触和体积锁定）。我们引入：i）层特异性功能，分离主动环向和纵向肌纤维，同时保持均匀的被动增强；ii）拟不可压缩体积贡献，以处理大的蠕动收缩。细胞电生理学进一步扩展，以再现慢波和尖峰爆发活动，从而首次在小肠段的三维几何中模拟局部神经兴奋。我们引入收缩力的时空调制，以准确捕获由慢波和尖峰爆发驱动的激活。整个耦合的非线性电力学边值问题遵循主动应变方法建模，还嵌入了鲁棒的增广拉格朗日接触算法，以避免大位移下的自穿透和几何不稳定性。然后使用GetFEM库中实现的内部P1-P2-P4有限元代码对8变量非线性控制方程进行离散化。数值实验表明，该框架能够再现生理蠕动，即壁收缩率大于80%，从而匹配体内内镜图像中的完全闭塞，并自然产生与实验观察一致的褶皱模式。我们表明，主动电力学各向异性异质建模策略对于数值稳定且生理准确的胃肠运动表征至关重要。
英文摘要
A new computational framework for modeling the intestinal wall as a multi-layered fiber-reinforced continuum is presented. The framework reproduces for the first time physiological motility and overcoming large-displacements limitations (self-contact and volume locking) occurring in classical hyperelastic formulations of soft tissues. We introduce: i) layer-specific functions, segregating active circumferential and longitudinal muscle fibers while maintaining homogeneous passive reinforcement, and ii) a quasi-incompressible volumetric contribution, to handle large peristaltic contractions. Cell electrophysiology is further extended to reproduce both slow waves and spike bursting activities thus mimicking for the first time a localized neural excitation in a three-dimensional geometry of small intestine segment. We introduce a spatio-temporal modulation of contractility to accurately capture activation driven by both slow waves and spike bursts. The overall coupled nonlinear electromechanical boundary valued problem is modeled following the active strain approach. A robust augmented-Lagrangian contact algorithm is also embedded to avoid self-penetration and geometrical instabilities under large displacements. The 8-variables nonlinear governing equations are then discretized using in house P1-P2-P4 finite elements codes implemented within the GetFEM library. Numerical experiments demonstrate the ability of the proposed framework to reproduce physiological peristalsis, i.e., wall contraction greater than 80%, thus allowing full occlusion matching in vivo endoscopic images, and naturally generating wrinkling patterns consistent with experimental observations. We show that an active electromechanics anisotropic heterogeneous modeling strategy is critical for a numerically stable and physiologically accurate representation of gastrointestinal motility.
