TOWARD PATIENT SPECIFIC DENSITY MODELLING OF THE TIBIA

这个子项目是许多研究子项目中利用 资源由NIH/NCRR资助的中心拨款提供。子项目和 调查员(PI)可能从NIH的另一个来源获得了主要资金， 并因此可以在其他清晰的条目中表示。列出的机构是 该中心不一定是调查人员的机构。 (我们要求加快审查。我和我的研究生Charles Negus一起进行了这项工作，在休息期间，我们想再做一次，以期待一个新的项目。摘要如下：虽然机械生物学骨骼研究始于细胞水平或低于细胞水平，但它最终应用于改善临床环境中的患者护理。弥合细胞水平研究和临床应用之间的差距的一个关键研究领域是计算力学。应力断裂形成是一种机械介导的病理学，它有可能直接得到计算方法和机械生物学研究的进步的帮助。应力性骨折病因学似乎是由个体生物力学、骨几何和骨质量引起的高度患者特有的骨负荷的相互作用。骨骼的几何形状和质量将内在地反映遗传影响和先前的功能适应。重复加载到骨骼上会导致疲劳样损伤累积，特别是在高度矿化的脆性骨骼中，在重复加载下，小裂纹会聚为应力性骨折[1]。骨折周围的局部重塑试图清除损伤，如果在持续加载的情况下发生，实际上可能会加剧问题。因此，评估个人的应力性骨折潜力需要评估他们的骨骼质量以及在重复载荷下的最大应力。评估受试者的胫骨应力分布建议进行针对患者的有限元分析，其几何形状理想地基于低辐射剂量、非侵入性成像技术。将骨小梁和皮质密度分布分配给该模型还存在其他挑战。我们之前已经描述了一种新的、但计算简单的方法来实现密度重分布和正交异性材料排列[2]。动态亚弹性重塑的灵感来自于目前对细胞力学转导的理解，认为细胞力学转导是基于动态负荷引起的间质流体流动。重塑算法被合并到使用消息传递接口并行的显式动态有限元代码中。重塑发生在离散的、周期性的时间步长上。基于局部应变率更新皮质和骨小梁的表观密度，并且由局部应力张量驱动主要材料方向的重新排列。应力的计算采用与路径相关的亚弹性本构关系，用Jaumann应力率表示。加载条件在正常步态和爬楼梯过程中应用峰值速率的瞬间被线性化，假设以足够的频率发生以超过细胞重塑激活的阈值周期数。涉及股骨近端的模拟结果表明，该动态方法的目标应变率为|Di|=1.7%/秒。此外，我们将DHR应用于患者特定的胫骨3D模型。收集了女性受试者左侧胫骨长度的4%、38%和66%的外周定量计算机断层扫描(PQCT)。基于这三次扫描，通过外推估计出整个胫骨表面的近似3D模型。每个特定主题的胫骨模型都被分配了标准化胫骨项目中的通用骨骺，并根据她的人体测量学进行了缩放。然后将pQCT扫描的骨膜边界导入骨干区域，并在骨痂和pQCT边界之间生成三维放样曲面。然后用三线性六面体单元对三维曲面进行网格划分。假设胫骨远端是固定的，根据每个受试者的体重定义五种倾斜加载条件，应用于内侧和外侧关节表面。DHR预测的骨干密度分布将与原始的pQCT扫描进行比较，以评估预测的皮质厚度与实际厚度。然后可以使用这些模型进行特定于患者的应力分析，这有助于评估可能导致应力性骨折的峰值应力。这项研究说明了现代细胞研究、成像技术和计算方法如何能够以一种具有潜在临床实用性的方式整合在一起。

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. (We are requesting an expedited review. I conducted this work with my graduate student, Charles Negus, and during the break, we would like to do one more run in anticipation of a new project. The abstract is below: While mechanobiological bone research begins at or below the cellular level, it finds its ultimate application in improving patient care in a clinical setting. A crucial field of study in bridging the gap between the cellular level research and clinical application is computational mechanics. Stress fracture formation is a mechanically mediated pathology that has the potential to be directly aided by advances in computational methods and mechanobiological research. Stress fracture etiology seems to be a highly patient-specific interplay of bone loads arising from an individuals biomechanics, bone geometry, and bone quality. Bone geometry and quality will inherently reflect both genetic influences and prior functional adaptation. Repeated loads to the bone result in fatigue-like damage accumulation, particularly in highly mineralized, brittle bone where small cracks coalesce into stress fractures under repeated loading [1]. Local remodeling around the fracture which attempts to clear the damage, if it occurs in the presence of continued loading, can actually exacerbate the problem. Assessing an individuals stress fracture potential thus entails assessing their bone quality as well as their peak stresses under repeated loads. Estimating a subjects tibial stress profile suggests a patient-specific finite element analysis, whose geometry is ideally based on low radiation dosage, noninvasive imaging techniques. Assigning trabecular and cortical density distributions to this model present other challenges. We have previously described a novel, but computationally simple approach to density redistribution and orthotropic material alignment [2]. Dynamic Hypoelastic Remodeling is inspired by current understanding of cellular mechanotransduction as being predicated on interstitial fluid flow resulting from dynamic loading. The remodeling algorithm is incorporated in an explicit, dynamic finite element code parallelized using Message Passing Interface. Remodeling occurs at discrete, periodically occurring time steps. Cortical and trabecular apparent densities are updated based on local strain rate, and realignment of principal material directions is driven by the local stress tensor. Stress is calculated using a path-dependent hypoelastic constitutive law, formulated with the Jaumann stress rate. Loading conditions were linearized around instants of peak rate of application during normal gait and stair climbing, assumed to occur with sufficient frequency to exceed the threshold cycle number for cellular remodeling activation. Results from simulations involving the proximal femur indicate that a target strain rate for this dynamic approach is |D_I | = 1.7%/sec. Additionally, we apply DHR to patient-specific 3D models of the tibia. Peripheral quantitative computed tomography (pQCT) scans of the left tibia were collected for female subjects at 4%, 38%, and 66% of tibial length. Based on these three scans, approximate 3D models of the entire tibial surface was estimated by extrapolation. Each subject-specific model tibia was assigned generic epiphyses from the Standardized Tibia Project which were scaled according to her anthropometrics. Then the periosteal boundaries of the pQCT scans were imported into the diaphyseal region, and a 3D lofted surface was generated between the epiphyses and pQCT boundaries. The 3D surface was then meshed with trilinear hexahedral elements. The distal end of the tibia was assumed fixed, and five ramped loading conditions, defined based on each subjects body weight, were applied to the medial and lateral articular surfaces. The diaphyseal density distribution predicted by DHR will be compared against the original pQCT scans to assess the predicted cortical thickness against actual thickness. These models can then be used to conduct patient-specific stress analyses which could aid in assessing peak stresses that could lead to stress fracture. This research illustrates how modern cellular research, imaging techniques, and computational methods can be integrated in a manner which has potential clinical practicality.