5 Modelling of the Brain for Injury Simulation and Prevention
123
measurement of biomechanical parameters, investigation of behavioural changes,
and evaluation of histological changes to identify the type and severity of brain
injury due to impact. More importantly, laboratory-raised animals are genetically
the same and can be well controlled for testing, once the experimental protocol
passes the stringent protocol review process.
Although experiments involving ‘volunteers’ (such as a collegiate football
player) provide valuable data for predicting the risk of brain injury, methods of
observing injuries within a live human brain are limited. To this extent, conducting
animal brain injury experiments and creating models based on these experiments
become critical. For studies on animal brains, we can compare brain regions with
high MPS, as predicted by a computational animal head model, to brain regions with
damage, as examined thoroughly using various histological methods. If the model
predictions and histology observations match, then we can conclude that the MPS
is a valid injury predictor, and if they do not match, new injury predictors, such as
pressure, strain rate, or others, could be evaluated. For example, a high-definition
3D rat head model with a detailed hippocampus, including the CA1, CA2, CA3,
and dentate gyrus (DG), has been used to demonstrate that high MPS is correlated
to region-specific hippocampal cell death [63] when subjected to equi-biaxial
strain fields. Similarly, Fig. 5.12 shows a simulated 3D rat brain responses under
controlled cortical impact. It can be seen that high-strain areas were concentrated
not only under the impactor but also away from the impactor near the occipital bone
when there were additional craniotomies performed (indicated as red-shaded circles
in Fig. 5.12) on the occipital bone. In this simple, straightforward experimental
model, the injury mechanism and associated injury threshold can be developed with
higher confidence than that from a concussed football player. Studies of this type
will help in developing more accurate region-specific injury assessment functions
that will help to better predict human head injury through human head models.
Even though animals are not human, if one cannot model the animal brain with
high degree of confidence, he/she is certainly not able to model the human brain
where experimental data are very hard to secure.
It has been nearly 20 years since animals larger than rodents were used to
determine TBI mechanisms and thresholds due to blunt impact. While valuable
data were gathered from animals in the old days, biomechanical parameters
measured were limited to mostly external impact parameters (such as the pendulum
mass and speed) and global responses (such as linear and angular accelerations).
Unfortunately, computer modelling technologies were not sophisticated enough to
model these experiments at that time, and hence anthropometry and other data
needed for validation of numerical models were not obtained. Nevertheless, full
utilisation of these data should be attempted so that no additional animals need to
be sacrificed unless it is absolutely necessary. It is our understanding that the US
Department of Transportation (DOT) is organising an effort to revive some primate
data from its archives across the globe. Successful retrieval of these primate data in
conjunction with numerical modelling of these animal impacts will promote a better
understanding of the relationship between injury outcome and model-predicted
intracranial responses.
123
measurement of biomechanical parameters, investigation of behavioural changes,
and evaluation of histological changes to identify the type and severity of brain
injury due to impact. More importantly, laboratory-raised animals are genetically
the same and can be well controlled for testing, once the experimental protocol
passes the stringent protocol review process.
Although experiments involving ‘volunteers’ (such as a collegiate football
player) provide valuable data for predicting the risk of brain injury, methods of
observing injuries within a live human brain are limited. To this extent, conducting
animal brain injury experiments and creating models based on these experiments
become critical. For studies on animal brains, we can compare brain regions with
high MPS, as predicted by a computational animal head model, to brain regions with
damage, as examined thoroughly using various histological methods. If the model
predictions and histology observations match, then we can conclude that the MPS
is a valid injury predictor, and if they do not match, new injury predictors, such as
pressure, strain rate, or others, could be evaluated. For example, a high-definition
3D rat head model with a detailed hippocampus, including the CA1, CA2, CA3,
and dentate gyrus (DG), has been used to demonstrate that high MPS is correlated
to region-specific hippocampal cell death [63] when subjected to equi-biaxial
strain fields. Similarly, Fig. 5.12 shows a simulated 3D rat brain responses under
controlled cortical impact. It can be seen that high-strain areas were concentrated
not only under the impactor but also away from the impactor near the occipital bone
when there were additional craniotomies performed (indicated as red-shaded circles
in Fig. 5.12) on the occipital bone. In this simple, straightforward experimental
model, the injury mechanism and associated injury threshold can be developed with
higher confidence than that from a concussed football player. Studies of this type
will help in developing more accurate region-specific injury assessment functions
that will help to better predict human head injury through human head models.
Even though animals are not human, if one cannot model the animal brain with
high degree of confidence, he/she is certainly not able to model the human brain
where experimental data are very hard to secure.
It has been nearly 20 years since animals larger than rodents were used to
determine TBI mechanisms and thresholds due to blunt impact. While valuable
data were gathered from animals in the old days, biomechanical parameters
measured were limited to mostly external impact parameters (such as the pendulum
mass and speed) and global responses (such as linear and angular accelerations).
Unfortunately, computer modelling technologies were not sophisticated enough to
model these experiments at that time, and hence anthropometry and other data
needed for validation of numerical models were not obtained. Nevertheless, full
utilisation of these data should be attempted so that no additional animals need to
be sacrificed unless it is absolutely necessary. It is our understanding that the US
Department of Transportation (DOT) is organising an effort to revive some primate
data from its archives across the globe. Successful retrieval of these primate data in
conjunction with numerical modelling of these animal impacts will promote a better
understanding of the relationship between injury outcome and model-predicted
intracranial responses.