Development and Validation of an MRI-Derived Head-Neck Finite Element Model

Introduction:: Throughout history, researchers have used computational head models based on the finite element (FE) method to investigate how the head responds to various types of injuries, such as those caused by contact sports, motor vehicle accidents, and violent impacts. These models have proven to be valuable tools in the field of biomechanics, providing insight into the mechanisms of traumatic brain injury and aiding in the development of protective equipment. Nonetheless, the majority of the existing finite element (FE)-based head and head-neck computational models suffer from biofidelic geometric and anatomical characterizations due mainly to coarse meshing, the lack of cerebrospinal fluid (CSF), neck muscles and ligaments, and simplistic material properties. Therefore, for a comprehensive understanding of the head and brain injury mechanism, this study aimed to develop and validate a biofidelic head-neck model.


Materials and Methods:: This study utilized magnetic resonance imaging (MRI) of a male participant (age: 42 years, height: 176 cm, and weight: 106 kg) to produce an accurate anatomical representation of the head and neck structure. The ' 'model's alignment was then corrected using principal component analysis to ensure a neutral supine position. We used the ANSA software to generate a high-quality mesh for the model, with 1.5 million elements. The different components of the head-neck were modeled using various material properties. For example, bones, skull, vertebrae, dura mater, and pia mater were modeled as linear elastic, while the scalp was modeled as linear viscoelastic. The brain was modeled as hyper-viscoelastic, the cerebral spinal fluid (CSF) as hyper-elastic, and the intervertebral discs as hyper-viscoelastic. The model also included 22 ligaments and 42 neck muscles, modeled as linear elastic beam elements and 1-D Hill-type beam elements, respectively.

We used LS-DYNA explicit solver to simulate the model, producing results such as normal stress, von Mises stress, maximum principal strain, and neck angle. We assessed the model's validity by replicating three experimental studies, including NBDL's high acceleration profile, Zhang's linear and rotational acceleration profiles, and Nahum's impact study. The accuracy of the model was evaluated by comparing the simulation results with the experimental data.


Results, Conclusions, and Discussions:: The head and neck geometry data exhibited good agreement with previous literature. The head-only model results were within the range of Zhang's numerical results, whereas the head-neck model showed a substantial reduction in shear stresses in midbrain and thalamus regions and intracranial pressure values. The kinematic responses of the head-neck model were accurate, with a strong positive correlation (r > 0.97) between experimental and numerical results. The intracranial pressure predictions for the head-only model showed good agreement with Nahum's experimental results, whereas the head-neck model exhibited decreased intracranial pressures.

The numerical results demonstrated that head-only and head-neck models accurately simulated the experimental impact scenarios, with the addition of neck structures significantly reducing intracranial pressure. We also observed a phase difference in the peak time of intracranial pressure between head-only and head-neck models during acceleration scenarios, indicating the importance of including neck structures in injury models. Additionally, replicating Nahum's study revealed a sharp increase in intracranial pressure values at the parietal and occipital lobes in the last quarter of the impact duration. This finding suggests that the already deformed neck structures fail to attenuate additional impact energy from the impactor in head-only models, highlighting the importance of including neck structures for a comprehensive understanding of head injury mechanics. Overall, these results underscore the importance of including neck structures in head injury models to improve injury prediction and prevention strategies. However, the model has a few limitations that need to be acknowledged, such as the brain white and gray matter and cancellous and cortical bones of skull and vertebra were not separately modeled, and extensor and flexor neck muscles were modeled using identical, constant activation levels. Lastly, the CSF was modeled as a solid with fluid-like behavior.

In this study, we developed and validated a novel, biofidelic head-neck FE model by using experimental data from the literature. The model can be a valuable tool for investigating head injury mechanisms and developing protective technologies against traumatic brain injury.


Acknowledgements (Optional): :

References (Optional): :