The Role of Diffusion Tensor
Imaging in Traumatic Brain Injury
Marilyn F. Kraus, MD, Associate Professor of Psychiatry
and Neurology, University of Illinois at Chicago and Researcher, Center
for Cognitive Medicine, mkraus@psych.uic.edu
Moderator: Kirby G. Vosburgh, PhD, Associate
Director, CIMIT and Assistant Professor, Radiology, Brigham and Women’s
Hospital, Harvard Medical School, kvosburgh@partners.org
Traumatic brain injury (TBI) is
a significant public health problem. Many people live with TBI-related
disabilities. More recently, it has been identified as a critical problem
in the military. Mild TBI is the most common type (also referred to
as concussion). Neurobehavioral deficits (changes in cognition, mood
and behavior) are very common causes of disability after TBI, and terms
such as post-concussion syndrome are used to describe these changes.
TBI is a multidisciplinary area that challenges artificial boundaries
within medical specialties, and research is needed to better understand
neuropathology and predictive factors in outcome.
Damage to the white matter tracts of the brain is a common result of
TBI, and may be the primary pathology in many cases. This injury is
known as diffuse axonal injury (DAI) or traumatic axonal injury (TAI).
Standard clinical neuroimaging is often not sensitive enough to detect
this. Diffusion tensor imaging (DTI) is one of the more recent tools
developed using MRI technology and allows for the specific examination
of the integrity of white matter tracts. Current research has shown
it to be sensitive to changes following even so-called mild TBI.
Mapping Brain Connectivity
with Diffusion MRI
Van Wedeen, MD, Assistant in Neuroscience, Massachusetts
General Hospital and Associate Professor in Radiology, Harvard Medical
School, van@nmr.mgh.harvard.edu
Moderator: Thomas J. Brady, MD,
Co-Program Leader, CIMIT Cardiovascular Disease Program; Director, Cardiovascular
Imaging and Intervention; Director, Nuclear Medicine and Molecular Imaging
and Vice Chairman, Director of Radiology Research, Massachusetts General
Hospital; L.L Robbins Professor of Radiology, Harvard Medical School,
tom@nmr.mgh.harvard.edu
Neuroanatomy is difficult because
brain structures overlap. Though the human brain may contain 10^14 synapses,
it is equally true that it consists of only a few hundred gray matter
“organs” –-cortical fields and subcortical nuclii -- connected
by a few times as many fiber pathways. Unfortunately for neuroscience,
the identity of theses structures has been difficult to determine. But
consider: if neural structures such as fiber pathways were spatially
separable, then like other anatomy, they would have been identified
by dissection and microscopy centuries earlier.
Diffusion spectrum MRI – DSI - and related methods offer a novel solution
to the problem of overlapping structure. DSI addresses the overlap of
fiber pathways by using the principle that the fibers of each pathway
are orientationally coherent – approximately parallel on the
scale of a voxel.
It follows that path anatomy naturally
resolved in a space of 5 dimensions – the 3 spatial dimensions plus
2 auxiliary dimensions to express fiber orientations at each location.
In retrospect, this solution to the problem of neoroanatomy was hiding
in plain sight. MRI and x-ray tomography failed to solve the problem
of neuroanatomy because each location of “white matter” and “gray
matter” in conventional tomograms typically includes many neural organs.
Yet just as the anatomic ambiguities of 2D projection x-ray are addressed
by 3D tomography, so the ambiguity of nueral structure in conventional
3D tomography may be resolved by DSI of auxiliary orientation dimensions.
As one application, Van Wedeen’s lab has recently shown that DSI can
produce a credible image of the human cortical ‘connectome’ –-
the totality of cortico-cortical connections in the human brain. With
further analysis, the lab is able to identify the topological core of
this network, showing it to be located principally on the medial surface
of the cerebral hemispheres in the poster cingulate /medial parietal
/ retro splenial region. This region, in fact, had been recently identified
by functional MRI including studies of the resting state as part of
the brain’s ‘default network’ - a core functional region active
when the brain is not attending to specific purposeful tasks. Quantitative
connectomics is anticipated to become a rich source of biomarkers for
a range of neurological and neuropsychiatric disease. Though still in
its infancy, diffusion MRI is emerging as a unique technology for neuroscience,
able to image the functional components of the brain quickly, objectively,
at low cost, in single brains in their entirety, nondestructively, ex
vivo and in vivo, in animals, and in human subjects.
Comments