Abstract:
This study aims to calculate the probabilistic response of a validated and verified parametric cervical spine finite element (FE) model, addressing the common limitation in computational analyses that fail to account for the inherent variability and uncertainty of biological systems. The model's geometry was defined using a set of parameters measured from Computed Tomography (CT) scans of 100 male and female volunteers. Material properties for the soft tissues of the cervical spine were derived from both experimental data and existing literature. This data was used to fit random distributions to both the geometric and material variables. The probabilistic analysis was performed using the NESSUS software, which propagates these input uncertainties through the FE model to predict a range of possible outcomes. For the analysis, a C5/C6 motion segment model was subjected to a 2 N-m moment in flexion and extension, and a mean value response was used to examine its probabilistic behavior. The results, compared with experimental data, demonstrated that even a relatively small coefficient of variation (C.O.V.) of 10% in the inputs could lead to significant variations in the model's response, particularly during flexion. The NESSUS software also provided importance values for each variable, indicating their influence on the model's response at different stages of loading. This methodology provides a powerful tool for predicting the probability of injury by accounting for biological variability, making it highly applicable in the field of biomechanics.
