In the past two decades, the awareness of the physical and emotional effects and
sequalae of traumatic brain injuries (TBI) has grown considerably, especially in
the case of soldiers returning from their deployment in Iraq and Afghanistan, after
sustaining blast-induced TBI (bTBI). While the understanding of bTBI and how it
compares to civilian non-blast TBI is essential for proper prevention, diagnosis and
treatment, it is currently limited, especially in human in-vivo studies.
Developing neuroimaging biomarkers of bTBI is key in understanding primary blast
injury mechanism. I therefore investigated the patterns of white matter and grey
matter injuries that are specific to bTBI and aren¶t commonl\ seen in civilians Zho
suffered from head trauma using advanced neuroimaging techniques. However,
because of significant methodological issues and limitations, I developed and
tested a new pipeline capable of running the analysis of white matter abnormalities
in soldiers, called subject-specific diffusion segmentation (SSDS). I also used
standard methodologies to investigate changes at the level of the grey matter
structures, and more particularly the limbic system. Finally, I trained a machine
learning algorithm that builds decision trees with the aim of classifying between
patients with TBI and controls, and between different TBI mechanisms as an
example of what could potentially be applied in the context of bTBI.
I found three main neuroimaging biomarkers specific to bTBI. The first one is a
microstructural white matter abnormality at the level of the middle cerebellar
peduncle, characterized by a decrease of diffusivity measures. The second is also
a decrease in diffusivity properties, at the level of the white matter boundary, and
the third one is a loss of hippocampal volume, with no association to post-traumatic
stress disorder. Finally, I demonstrated that SSDS can be used in tandem with a
machine learning algorithm for potential diagnosis of TBI with high accuracy.
These findings provide mechanistic insights into bTBI and the effect of primary blast
injuries on the human brain. This work also identifies important neuroimaging
biomarkers that might facilitate prevention and diagnosis in soldiers who suffered from
bTBI.