Vertebral strength prediction under anterior compressive force using a finite element model for osteoporosis assessment

Eleven lumbar spines (5 M and 6F aged: 82 years ± 7) were scanned on a qCT machine (Scanner ICT 256, Philips Healthcare, Cleveland, OH, 120 kV, 1489 mA/s and a voxel size: 0.39 mm × 0.39 mm × 0.33 mm) along with a calibration phantom (QRM-ESP, QRM GmbH, Germany) to map gray scale values to BMD. DXA measurements (Hologic Inc, Waltham, MA, USA) were performed with each spine positioned in a 15 cm water bath. BMD was calculated using the A-P scanning protocol. The 28 vertebrae (8 L1, 11 L2 and 9 L3) were then cleaned from all soft tissue and the posterior elements were transected. The vertebral bodies were potted in PMMA for parallelism before anterior compressive tests were conducted using a spherical seating loading platen (Instron Ltd., High Wycombe, UK) with the centre of rotation aligned with the anterior third of the vertebral body (Figure 1). Specimens were destructively tested in compression at 1 mm/min. Vertebral strength was defined as the ultimate load achieved and axial stiffness was calculated as the slope of the force–displacement curve. A FEM was built based on the qCT images using a hexahedral mesh generated with a custom-built algorithm (8300 elements and 190,306 nodes). Material properties for each element were assigned using density-modulus relationship (Kopperdahl et al. 2002). PMMA was modelled (E = 2.5 GPa, ν = 0.3) with the lower layer constrained and the upper layer joined by rigid elements to a node located at the anterior third of the vertebra to apply anterior compressive load. TMSimulations were run on ANSYS software (ANSYS Inc., Canonsburg, PA, USA). The vertebral failure load was Vertebral strength prediction under anterior compressive force using a finite element model for osteoporosis assessment