Where
did we start?
We
first became interested in the mechanics of traumatic brain injury several years
ago, where we studied some experimental models of brain injury. Dave
Shreiber, a lab alumnus now on the faculty in biomedical engineering at
Rutgers University, used computational mechanics tools to predict the mechanical
conditions that corresponded to injuring brain tissue. The estimates of the
tissue tolerance have been helpful in developing computational models of the
human brain response during motor vehicle accidents, an activity that is now
being used by the federal government to see if they can develop a better way
to predict the risk of head injuries in car accidents (see
this link for more details).
We first started looking at this
problem by examining the axonal projection of neurons, since there was past
information that showed these cellular elements were consistently injured
in almost all head injured patients. However, the pattern of axonal damage
in these patients is not uniform throughout the brain, and it does not affect
all axons at a specific point within the brain. Remarkably, damaged or swollen
axons are almost always completely surrounded by normal, healthy looking axons
(see figure at left).
We measured the shape of axons
in white matter and predicted how the shapes would change when the tissue
was deformed. We then compared the predictions to the measured shape changes
that occurred when we stretched samples of white matter tissue. We found that,
at low levels of stretch deformation, the individual axons would elongate
without any interactions with neighboring axons. At higher levels of deformation,
though, the axons became coupled and interlinked. The cells that probably
provide this coupling and 'cross-linking' are the oligodendrocytes, which
myelinate several axons in parallel within the white matter. These findings
were important, since it provided a partial explanation of why so few axons
show damage in traumatic brain injury. It also explains why white matter tissue
has such nonlinear material properties. (For more detail, you can look at
some papers that we recently published on this work - paper
1, paper 2, paper
3).
Research
Areas
Cross-section
of white matter tissue in the brain, stained with an antibody to detect the
neurofilament proteins within the axon. Large acumulations of the neurofilament
protein indicate axons that are damaged. Even though the injury was severe,
there are remarkably few axons that show damage.
These
computational tools did not allow us to predict, though, how individual neurons
were experiencing the macroscopic forces applied to the tissue. We did not know
if the neurons in a given region of the brain experienced the same mechanical
force as their neighbor, or if there was a tremendous variation in the forces
applied at the cellular level. We felt that knowing how forces 'transferred'
to individual cellular elements would be key in deciding how populations of
neurons within a brain region could receive a very broad distribution of mechanical
forces at the moment of brain injury.
