Experimental Testing and Biomechanical CT Analysis of Chinese Cadaveric Vertebrae with Different Modeling Approaches
Yi Wei,Wentian Feng,Guanghui Li,Zuchang Li,Zaiwei Liu,Xiaoguang Cheng,Haisheng Yang
DOI: https://doi.org/10.1016/j.medengphy.2021.05.008
IF: 2.356
2021-01-01
Medical Engineering & Physics
Abstract:Osteoporosis is characterized by reduced bone strength predisposing to an increased risk of fracture. Biomechanical computed tomography (BCT), predicting bone strength via CT-based finite element analysis (FEA), is now clinically available in the USA for diagnosing osteoporosis or assessing fracture risk. How-ever, it has not been previously validated using a cohort of only Chinese subjects. Additionally, the effect of various modeling approaches on BCT outcomes remains elusive. To address these issues, we performed DXA and QCT scanning, compression testing, and BCT analyses on thirteen vertebrae derived from Chi-nese donors. Three BCT models were created (voxBCT and tetBCT: voxel-based and tetrahedral element -based FE models generated by a commercial software; matBCT: tetrahedral element-based FE model generated by a custom MATLAB program). BCT-computed outcomes were compared with experimental measures or between different BCT models. Results showed that, DXA-measured areal bone mineral den-sity (aBMD) showed weak correlations with experimentally-measured vertebral stiffness (R-2 = 0.28) and strength (R-2 = 0.34). Compared to DXA-aBMD, BCT-computed stiffness provided improved correlations with experimentally-measured stiffness (voxBCT: R-2 = 0.82; tetBCT: R 2 = 0.77; matBCT: R-2 = 0.76) and strength (voxBCT: R-2 = 0.55; tetBCT: R-2 = 0.57; matBCT: R-2 = 0.53); BCT-computed mechanical param-eters (stiffness, stress and strain) of the three different models were highly correlated with each other, with coefficient of determination (R-2) values of 0.89-0.98. These results, based on a cohort of Chinese vertebral cadavers, suggest that BCT is superior over aBMD to consistently predict vertebral mechanical characteristics, regardless of the modeling approaches of choice. (C) 2021 IPEM. Published by Elsevier Ltd. All rights reserved.