Image Processing Reference
In-Depth Information
23.2.3 Multi-field Data
Diffusion Tensor Imaging, or DTI, is an MRI-based acquisition modality, introduced
in 1994 by Basser et al., that yields 3
3 symmetric diffusion tensors as its native
measurement quantity [
5
]. The tensors represent the local diffusion of water mole-
cules, and hence indirectly indicate the presence and orientation of fibrous structures,
such as neural fiber bundles or muscle fibers. Already in this first paper, the authors
employed 3D glyphs to visualize the eigensystems represented by the tensors.
Basser and his colleagues were also some of the first to extract and visualize fiber-
tract trajectories from DTI data of the brain [
4
,
6
], thus linking together the point
diffusion measurements to get an impression of the global connective structures in
the brain. With DTI it was encouraging to see that the first visualization efforts were
initiated by the scientists developing this new scanning modality themselves. Early
work by the visualization community includes tensor lines for tractography [
76
] and
direct volume rendering of DTI data [
38
,
39
].
Importantly, DTI serves as one of the first examples of nativelymulti-fieldmedical
data, that is medical data with multiple parameters defined over the same spatio-
temporal domain. The advent of DTI initiated a whole body of medical visualization
research dedicated to the question of how best to visually represent and interact with
diffusion tensor data in particular and multi-field medical data in general. The 2007
paper by Blaas et al. presented a visual analysis-inspired solution to this problem
based on linked physical and feature space views [
9
].
×
23.2.4 Time-Varying Data
Time-varyingmedical volume data visualizationmade its entrance in 1996 with work
by Behrens et al. [
7
] on supporting the examination of Dynamic Contrast-Enhanced
MRI mammography data with the display of parameter maps, the selection of regions
of interest (ROIs), the calculation of time-intensity curves (TICs), and the quantitative
analysis of these curves. In 2001, Tory et al. [
69
] presented methods for visualizing
multi-timepoint (1 month interval) MRI data of a multiple sclerosis (MS) patient,
where the goal was to study the evolution of brain white matter lesions over time.
Methods used included glyphs, multiple isosurfaces, direct volume rendering and
animation. Coto et al. [
16
] applied multiple coupled views, including linked cursors
and brushing on 3D renderings and scatterplots, to dynamic contrast-enhanced MRI
(DCE-MRI) mammography data.
23.2.5 Illustrative Visualization
Illustrative visualization is primarily motivated by the attempt to create renditions
that consider the
perceptual capabilities
of humans. As an example, humans infer
Search WWH ::
Custom Search