Fuzzy Fibers: Uncertainty in dMRI Tractography - Scientific Visualization: Uncertainty, Multifield, Biomedical, and Scalable Visualization - page 82

SH coefficients of the voxels with the highest FA, presumably representing a single
fiber.
Aside from the question of counting fibers, other work has examinedmore broadly
the question of which models of the diffusion weighted signal profile are statistically
supported. Bretthorst et al. [ 8 ] compute Bayesian evidence (see Sect. 8.4.1 ) to quan-
tify the fitness of various models of the diffusion weighted signal, producing maps
of model complexity in a fixed baboon brain, and of evidence-weighted averages of
per-model anisotropy. Freidlin et al. [ 17 ] choose between the full diffusion tensor and
simpler constrained tensor models according to the Bayesian information criterion
(BIC) or sequential application of the F-Test and either the t-Test or another F-Test.
8.3.3 Partial Voluming
Tractography works best in voxels that contain homogeneously oriented tissue.
Unfortunately, many regions of the brain exhibit more complex structures, where
fibers cross, diverge, or differently oriented fibers pass through the same voxel [ 1 ].
This problem is reduced at higher magnetic field strength, which affords increased
spatial resolution. However, even at the limit of what is technically possible today
[ 20 ], a gap of several orders of magnitude remains to the scale of individual axons.
Super-resolution techniques combine multiple images to increase effective reso-
lution. Most such techniques use input images that are slightly shifted with respect
to each other and initial success has been reported with transferring this idea to MRI
[ 47 ]. However, due to the fact that MR images are typically acquired in Fourier
space, spatial shifts do not correspond to a change in the physical measurement, so
it is unclear by which mechanism repeated measurements should achieve more than
an improved signal-to-noise ratio [ 48 , 51 ]. It is less controversial to compute images
that are super-resolved in slice-select direction [ 18 , 52 ] or to estimate fiber model
parameters at increased resolution via smoothness constraints [ 42 ].
Track density imaging [ 10 ] uses tractography to create super-resolved images
from diffusion MRI. After randomly seeding a large number of fibers, the local
streamline density is visualized. It is computed by counting the number of lines
that run through each element of a voxel grid whose resolution can be much higher
than during MR acquisition. Visually, the results resemble those of line integral
convolution, which had been applied to dMRI early on [ 23 , 69 ].