股骨颈有限元分析的赋材料属性方法探讨及有效性验证

Methods for material assignment of femoral neck finite element analysis and its valid confirmation

背景:作者前期实验通过理论推导的方式,认为将骨骼材料属性分为10种即可达到有限元分析的要求,其结论是否与真实相符,有待于与体外力学实验结果进行对比验证。目的:对12侧股骨上段标本进行有限元分析,并与体外力学试验结果对比验证,以探讨骨骼合理的赋材料属性方法。方法:将12侧股骨上段标本进行高速CT薄层扫描,在Mimics中进行三维重建,在Ansys中进行体网格划分。有限元分析组在Mimics中根据灰度值将三维模型材料属性分为2种(密质骨及松质骨)、10,50,100,200,400种后赋予材料属性,在Ansys中进行力学分析并采集股骨颈表面节点力学数值。体外力学实验组将12侧股骨上段标本进行压缩实验,采集与有限元分析组相同的测试点力学数据。结果与结论:单因素方差分析显示,力学实验组与有限元分析2种材料属性组比较,差异无显著性意义(P=0.082);与有限元分析10,50,100,200,400种材料属性组比较,差异无显著性意义(P>0.39)。有限元分析各亚组中,2种材料属性组与其余各组比较,差异有显著性意义(P<0.05),10,50,100,200,400种材料属性组间比较,差异无显著性意义(P>0.9)。结果提示,赋予骨骼10种材料属性即可达到有限元分析的要求。

BACKGROUND： Based on previously theoretical derivation, it thought that assignment with 10 kinds of material attributes to three-dimensional model of bone can match the needs of finite element analysis, however, whether the results is consistent with actual needs to be validated by experimental results. OBJECTIVE： Twelve specimens of femoral superior segment were used for finite element analysis, which were verified with results of biomechanical testing, to explore a reasonable method for material assignment of bone. METHODS： All 12 specimens of femoral superior segment were treated with CT scan, three-dimensional reconstruction in Mimics 10.0 and volume meshing in Ansys. The finite element analysis group was divided into 2 kinds （compact bone and cancellated bone）, 10, 50, 100, 200, 400 kinds of material attributes groups based on the gray value. All models were assigned with material attributes and tested in Ansys for mechanics data of nodes on surface of femoral neck. In biomechanical testing group, 12 specimens of femoral superior segment were treated with compressed testing to harvest mechanics data of measuring point same as that of finite element analysis group. RESULTS AND CONCLUSION： The one-way analysis of variance showed that the differences between the biomechanical testing group and finite element analysis group of 2 kinds of material attributes had no obvious significance （P=0.082）. Compared to the 10, 50, 100, 200, 400 kinds of material attributes group, the difference had no significance （P〉 0.39）. However, the differences between the 2 kinds of material attributes and the 10, 50, 100, 200, 400 kinds of material attributes in the finite element analysis group were obviously difference （P 〈0.05）, which was no difference in the 10, 50, 100, 200, 400 kinds of material attributes （P〉 0.9）. The results demonstrated that to assign 10 kinds of material attributes to three-dimensional model of bone can match the needs of finite element analysis.