Comparing FE human body model rib geometry to population data
Sven A. Holcombe,Amanda M. Agnew,Brian Derstine,Stewart C. Wang
DOI: https://doi.org/10.1007/s10237-020-01335-2
IF: 3.5
2020-05-22
Biomechanics and Modeling in Mechanobiology
Abstract:Finite element human body models (HBMs) are used to assess injury risk in a variety of impact scenarios. The ribs are a key structural component within the chest, so their accuracy within HBMs is vitally important for modeling human biomechanics. We assessed the geometric correspondence between the ribs defined within five widely used HBMs and measures drawn from population-wide studies of rib geometry, focusing on (1) rib global shape, (2) rib cross-sectional size and shape, and (3) rib cortical bone thickness. A parametric global shape model fitted to all HBM ribs was compared to expected rib parameters calculated for each HBM’s subject demographic using population reference data. The GHBMC M50 and THUMS M50 male HBMs showed 24% and 50% of their fitted rib shape parameters (6 parameters per each 12 ribs) falling outside 1SD from population expected values, respectively. For female models the GHBMC F05, THUMS F05, and VIVA F50 models had 21%, 26%, and 19% of their rib shape parameters falling outside 1SD, respectively. Cross-sectional areas and inertial moments obtained along the HBM ribs were compared to average ± 1SD corridors for male and female ribs drawn from reference population data. The GHBMC M50, THUMS M50, and VIVA F50 model ribs were all larger in overall cross-sectional area than their targeted average population values by 0.9SDs (average across the rib’s full length), 1.7SDs, and 1.3SDs, respectfully. When considering cortical bone cross-sectional area, the THUMS and VIVA models—which each define a constant bone thickness value across the entire rib—overestimated bone content on average by 1.1SDs and 1.2SDs, respectively. HBMs have traditionally performed poorly when predicting rib fracture onset or fracture site, and in all HBMs in this study the rib regions with the most extreme cortical bone thickness and cross-sectional area discrepancies (compared to average reference data) corresponded to regions toward the sternal end of the ribs where rib fractures most frequently occur. Results from this study highlight geometrical components of current HBM ribs that differ from the rib geometry that would be expected from within those models’ target demographics, and help researchers prioritize improvements to their biofidelity.
engineering, biomedical,biophysics