Weinans H, Sumner D R, Igloria R, Natarajan R N
Erasmus Orthopaedic Research Laboratory, Departments of Orthopaedics, Erasmus University EE1614, P.O. Box 1738, 3000, Rotterdam, The Netherlands.
J Biomech. 2000 Jul;33(7):809-17. doi: 10.1016/s0021-9290(00)00036-1.
Subject-specific finite element (FE) computer models of the proximal femur in hip replacement could potentially predict stress-shielding and subsequent bone loss in individual patients. Before such predictions can be made, it is important first to determine if between subject differences in stress-shielding are sensitive to poorly defined parameters such as the load and the bone material properties. In this study we investigate if subject-specific FE models provide consistent stress-shielding patterns in the bone, independent of the choice of the loading conditions and the bone density-modulus relationship used in the computer model. FE models of two right canine femurs with and without implants were constructed based on contiguous computed tomography (CT) scans so that subject-specific estimates of stress-shielding could be calculated. Four different loading conditions and two bone density-modulus relationships were tested. Stress-shielding was defined as the decrease of strain energy per gram bone mass in the femur with the implant in place relative to the intact femur. The analyses showed that for the four loading conditions and two bone density-modulus relationships the difference in stress-shielding between the two subjects was essentially constant (1% variation) when the same loading condition and density-modulus relationship was used for both subjects. The severity of stress-shielding within a subject was sensitive to these input parameters, varying up to 20% in specific regions with a change in loading conditions and up to 10% for a change in the assumed density-modulus relationship. We conclude that although the choice of input parameters can substantially affect stress-shielding in an individual, this choice had virtually no effect on the relative differences in femoral periprosthetic stress-shielding between individuals. Thus, while care should be taken in the interpretation of the absolute value of stress-shielding calculated with these type of models, subject-specific FE models may be useful for explaining the variation in bone adaptation responsiveness between different subjects in experimental or clinical studies.
髋关节置换术中股骨近端的个体特异性有限元（FE）计算机模型有可能预测个体患者的应力遮挡及随后的骨质流失。在进行此类预测之前，首先确定个体间应力遮挡差异是否对应力和骨材料特性等定义不明确的参数敏感非常重要。在本研究中，我们调查了个体特异性有限元模型是否能在骨中提供一致的应力遮挡模式，而与计算机模型中使用的加载条件选择和骨密度 - 模量关系无关。基于连续的计算机断层扫描（CT）构建了两只右犬股骨有无植入物的有限元模型，以便能够计算个体特异性的应力遮挡估计值。测试了四种不同的加载条件和两种骨密度 - 模量关系。应力遮挡定义为植入物在位时股骨中每克骨质量的应变能相对于完整股骨的降低。分析表明，对于四种加载条件和两种骨密度 - 模量关系，当两个个体使用相同的加载条件和密度 - 模量关系时，两个个体之间的应力遮挡差异基本恒定（变化1%）。个体内部应力遮挡的严重程度对这些输入参数敏感，在特定区域随着加载条件的变化变化高达20%，对于假定的密度 - 模量关系的变化变化高达10%。我们得出结论，虽然输入参数的选择可显著影响个体的应力遮挡，但这种选择对个体间股骨假体周围应力遮挡的相对差异几乎没有影响。因此，虽然在解释用这类模型计算出的应力遮挡绝对值时应谨慎，但个体特异性有限元模型可能有助于解释实验或临床研究中不同个体间骨适应性反应的差异。
