Effect of CT-based material grouping on finite element strength and stiffness predictions in vertebrae with metastatic lesions
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Introduction
Metastatic spinal lesions substantially alter vertebral mechanical properties and increase fracture risk. Computed tomography (CT)-based finite element (FE) models can estimate vertebral strength, but their accuracy depends on how CT-derived material properties are represented. This study evaluated the effect of two material-grouping strategies on simulated strength and stiffness in metastatic vertebrae.
Methods
We compared Adaptive Clustering (AC) with Uniform fixed-width grouping in 44 vertebrae from 11 donors (8 osteolytic, 12 osteoblastic, 12 mixed, 12 no observed lesion (NOL)). FE models were generated based on CT scans with 2–500 material groups and compared for material-mapping error and simulated strength and stiffness. Overall and lesion-stratified agreement with experimental measurements was assessed in an exploratory analysis.
Results
AC showed significantly lower Young’s modulus root-mean-square error than Uniform (p < 0.05). Simulated strength and stiffness stabilised by 50 material groups. At 50 groups, simulated strength showed moderate correlation with experimental strength overall (R² = 0.57), strongest in NOL vertebrae (R² = 0.82) and lower in lesion-bearing vertebrae (R² = 0.41–0.59). Stiffness showed weaker correlation overall (R² = 0.27), highest in NOL vertebrae (R² = 0.48) and negligible in mixed lesions (R² = 0.007). Bland-Altman analyses indicated systematic underestimation of experimental fracture load.
Discussion
AC improved material-mapping fidelity, whereas increasing material groups beyond 50 had little influence on simulated strength or stiffness. Numerical stabilisation therefore did not imply experimental accuracy. Lesion-stratified findings were exploratory and should be interpreted cautiously because of limited subgroup sizes.
