Chemistry Reference
In-Depth Information
Figure 22.
An inversion recovery
image contrasting GM and WM.
duration, or shape of the diffusion weighting gradients. Thus it should be possible
to vary any of the characteristics of the diffusion gradients and, if the appropriate
mathematical expression is applied, a similar value of the diffusion coefficient for
the same sample should be obtained.
The sequence designs for the experiments in human neuronal tissue were in-
tended to vary the diffusion weighting in a variety of ways, while employing simple
methods. All components of the sequences with the exception of diffusion gradi-
ent design were identical. Gray and white matter tissue was differentiated using
inversion recovery images. An example of an inversion recovery image is shown in
Fig. 22. An inversion recovery sequence provides strong contrast between tissues
with different
T
1
relaxation times, such as GM and WM.
In Figs. 23 - 28, maps of the fitted parameters (
D
and
α
) from the fit of the
fractional equation [Eq. (181)] are shown for every tissue voxel in an image slice.
The same slice in the same subject was chosen across each of the three experiments,
with variable duration, variable strength, and variable shaped diffusion weighting
gradients. Next to each map is a histogram displaying the frequency of occurrence
of particular values of the parameters, as a percentage of the total fitted voxels. The
fractional diffusion equations fitted the majority of the voxels containing tissue,
with a chi-squared goodness of fit of
<
0
.
001. The broadness of the distributions
of the fitted parameters for gray and white matter can be attributed to the relatively
crude tissue differentiation provided by the inversion recovery image. The image
voxels would contain both cerebrospinal fluid (CSF) and partial volume effects
(multiple tissue types in the same voxel). The mean values of the distributions
are collated in Table III. It was conjectured that the nonmonoexponential behavior
