ASSAF Yaniv DTI White Matter 11 Oct 2011
<p class="MsoNormal" >ASSAF Yaniv DTI White Matter 11 Oct 2011</p>
<p>
I sit with <b>Nama Barnea</b> who works with Allan Reiss in Psychiatry. she is a psychiatrist and basic researcher.
She and he answer my Q abt DTI in CA3 in HC. (answer is that there ARE anisotropic structures even in CA3 (eg astrocytic processes
and dendrites of neurons)</p>
<p>
intro by Brian Wandell</p>
<p>
Exploring White Matter with DTI</p>
<p>
3d EM of white matter... forest of fibers</p>
<p>
<b>brain and spaghetti</b></p>
<p>
<b>spaghetti = white matter model</b></p>
<p>
<b>cooked spaghetti (with meatballs) </b>limits on DTI</p>
<p>
has gray matter or csf
different directions of fibers
So, <b> measured tensor </b>is an average</p>
<p>
<b>most ordered tract is the corpus callosum</b></p>
<p>
<b>uses the b-value</b> formalism</p>
<p>
it assumes that the displacment prob func is gaussian.</p>
<p>
diffusion in brain is NOT gaussian</p>
<p>
he plots ln (E) against the b value</p>
<p>
linear up to <b>2500 seconds per mm2</b></p>
<p>
Dslow is prob intracellular (slow cuz of proteins)
D fast is extracellular
but Aslow = 17% and A fast = 80% proportions... does not make sense
cuz most water is intracellular...</p>
<p>
look at bovine optic nerve</p>
<p>
how to deal with restricted diffusion</p>
<p>
either non-parametric or parametic models</p>
<p>
he likes Charmed and AxCaliber</p>
<p>
measuring displacment (rather than diffusion)</p>
<p>
displacement of water molecules</p>
<p>
1990 indicator of compartmentation... q -space imaging</p>
<p>
if u scan perpendic to fiber then can read out the diameter..</p>
<p>
The Composite Hindered and Restricd Model of Diffusion = CHARMED</p>
<p>
look at corpus caloss... restricted diffusion is from axons but in extracel matrix it is hindered;</p>
<p>
restricted diffusion... h2o bouncing around inside of axon</p>
<p>
membrane permeability and exchange is important issue...</p>
<p>
in rbc's h2o easily moves in and out in 15 msec.</p>
<p>
bottom line takes water 1/2 second = 500 msec to get thru myelinated axon.... really slow!!!
cuz of all those lamellae...</p>
<p>
so, cell soma takes 15 msec
but when u include myelin... it is really hindered... 500 msec...</p>
<p>
CC has astrocytes and oligos that are creating anisotropy;</p>
<p>
b-value is the key variable in DTI
looking for high B value contrast.
NAMA does DTI in psychiatry;</p>
<p>
so, CHARMED does a good job of separating crossing fibers:
eg cingulum over the CC</p>
<p>
from CHARMED to AxCaliber</p>
<p>
can distinguish large and small axons by allowing more time for diffusion...
small axons are restricted with small times
but large are not
but all axons including large will be restriced if time is large enuf.</p>
<p>
2 microns is avg diameter in optic N
in sciatic nerve avg diamter is 20 microns... of axons...</p>
<p>
next look at rat CC</p>
<p>
the CC is the largest white matter struc in the brain</p>
<p>
he can determine in vivo the diameter of axon diameters...</p>
<p>
AxCaliber = axon caliber</p>
<p>
axons are abt 3 microns in body of CC and perhaps 1 micron in genu and in splenium</p>
<p>
APPLICATIONS: White Matter Plasticity</p>
<p>
Scholz Training induces changes in white matter archit Nature Neurosci (poss. the juggling paper)</p>
<p>
<b>Fornix connects HC to septal nuclei thru the fimbria</b>..</p>
<p>
recall that in L group = learning group u do see
a reduc in ADC but not in control group</p>
<p>
next look at Water Maze and Ax Caliber looking at CC
after<b> 5 days of water maze training</b></p>
<p>
no change in control but u do see a big change
in Learning group.... get signif more small axons (1 micron diameter)...
even see it after 1 day in water maze....</p>
<p>
can read out <b>axonal density and axon diamter</b>...</p>
<p>
basis of models is the cylindr shape of axons.</p>
<p>
CHARMED and AxCaliber...</p>
<p>
offers VIRTUAL HISTOLOGY... (usu only measurable by electron microscopy)</p>
<p>
Q correl bt length and diameter of axon?
</p>