5 Modelling of the Brain for Injury Simulation and Prevention
103
Clark et al. [9] at the University College Dublin used a combined experimental and numerical approach by inputting the experimentally obtained linear and
angular kinematic data to run the University College Dublin Brain Trauma Model
(UCDBTM) for assessing the injury risk on ice hockey goaltender. As part of the
Total Human Model for Safety (THUMS) continuing advancements at the Toyota
Central R&D Labs, Atsumi et al. [6] reported a head model consisted of detailed
descriptions of the deep brain (e.g. the basal ganglia and fornix) and incorporated
strain rate-dependent, anisotropic properties for the brain parenchyma. Additionally,
Miller et al. [69] at Wake Forest University developed a voxel-based human head
model from a brain atlas available at the International Consortium for Brain
Mapping (ICBM) and included gyral folds in this model, just like that recommended
by Ho and Kleiven [34]. Many of the recently developed head models were
quantitatively evaluated in accordance to the CORrelation and Analysis (CORA)
method, which combined the evaluation of extent of cross correlation and phase
shift and the degree of convergence between the simulation and experimental results
[17].
As more head/brain FE models are reported from around the world, it is noticed
that many models are repetitive of previous versions with limited incremental
advancements [104]. The same limited experimental datasets were treated as ‘gold
standard’ and repetitively used for model validation without ever questioning the
size differences among the experimental subjects and the head models developed.
As long as site-specific injury-based model validations are not achieved, any new
models will continue to provide limited insights into the solution of the head injury
problem. The purpose of this chapter is to outline what have learned in the past
several decades regarding the development of human and animal brain models and
how the knowledge gained in studying head injury biomechanics has helped us
approaching the goals of mitigating brain injury.
5.2 Challenges of Investigating TBI Using FE Head Models
Mechanical properties of multiphasic biological tissues, by nature, are nonlinear,
viscoelastic, and anisotropic. Additionally, properties of these tissues are likely to
be age-, gender-, species-, and site-specific. Until a large-scale experimental study
is conducted to clarify the rate-, direction-, gender-, and age-dependent constitutive
laws and associated properties, cautions are needed before trusting any models
developed thus far which can actually be used to accurately predict the extent and
location of TBI. Discussion for the various structures and materials within the brain
will not be covered in this chapter as more information are provided in Chaps. 2
and 4. Aside from this big issue related to proper identifications of mechanical
properties, many other challenges exist, and these issues are briefly discussed in
this section.
103
Clark et al. [9] at the University College Dublin used a combined experimental and numerical approach by inputting the experimentally obtained linear and
angular kinematic data to run the University College Dublin Brain Trauma Model
(UCDBTM) for assessing the injury risk on ice hockey goaltender. As part of the
Total Human Model for Safety (THUMS) continuing advancements at the Toyota
Central R&D Labs, Atsumi et al. [6] reported a head model consisted of detailed
descriptions of the deep brain (e.g. the basal ganglia and fornix) and incorporated
strain rate-dependent, anisotropic properties for the brain parenchyma. Additionally,
Miller et al. [69] at Wake Forest University developed a voxel-based human head
model from a brain atlas available at the International Consortium for Brain
Mapping (ICBM) and included gyral folds in this model, just like that recommended
by Ho and Kleiven [34]. Many of the recently developed head models were
quantitatively evaluated in accordance to the CORrelation and Analysis (CORA)
method, which combined the evaluation of extent of cross correlation and phase
shift and the degree of convergence between the simulation and experimental results
[17].
As more head/brain FE models are reported from around the world, it is noticed
that many models are repetitive of previous versions with limited incremental
advancements [104]. The same limited experimental datasets were treated as ‘gold
standard’ and repetitively used for model validation without ever questioning the
size differences among the experimental subjects and the head models developed.
As long as site-specific injury-based model validations are not achieved, any new
models will continue to provide limited insights into the solution of the head injury
problem. The purpose of this chapter is to outline what have learned in the past
several decades regarding the development of human and animal brain models and
how the knowledge gained in studying head injury biomechanics has helped us
approaching the goals of mitigating brain injury.
5.2 Challenges of Investigating TBI Using FE Head Models
Mechanical properties of multiphasic biological tissues, by nature, are nonlinear,
viscoelastic, and anisotropic. Additionally, properties of these tissues are likely to
be age-, gender-, species-, and site-specific. Until a large-scale experimental study
is conducted to clarify the rate-, direction-, gender-, and age-dependent constitutive
laws and associated properties, cautions are needed before trusting any models
developed thus far which can actually be used to accurately predict the extent and
location of TBI. Discussion for the various structures and materials within the brain
will not be covered in this chapter as more information are provided in Chaps. 2
and 4. Aside from this big issue related to proper identifications of mechanical
properties, many other challenges exist, and these issues are briefly discussed in
this section.