Biomedical Engineering Reference
In-Depth Information
(a)
(b)
P
C
P
P
Q
R
P
Q
Q
Q
R
R
C
R
FIGURE 7.7
Plan (a) and profiles (b) of two plateaus and a border zone showing signal intensity
gradients in different directions. The profiles (PP, QQ, and RR) show the signal intensity
changes in passing from one plateau across the border zone to the other plateau. The
contour (CC) marks the edge of the high signal plateau. At each point on CC the
maximum signal intensity gradient is perpendicular to the contour. The direction of
the gradient defined as being from high to lower signal intensity is indicated by the
arrows. An estimate of the slope of the signal intensity gradient through PP, QQ, and
RR can be obtained by noting that the slope is inversely related to the distance between
the two plateaus.
7.4.2.4
Effects of Size of Shift
As shown in Figure 7.8 the initial effect of a shift is to show a small change on
the difference image (i.e., a border zone shift). If the shift is increased, a peak
is reached when one plateau just crosses the boundary zone to reach the other
plateau. This peak is maintained in intensity and widens as the shift is
increased further (i.e., an interplateau shift). The area under the curve on the
difference image increases in direct proportion to the size of the shift (i.e.,
area of a parallelogram) and is independent of the magnitude of the signal
intensity gradient. Using the plateau model, the size of the component of a
shift parallel to the local gradient direction will be the width shown on the
difference image multiplied by the mean difference signal as a fraction of the
full scale signal intensity.
7.4.2.5
Effects of Shift Direction
Shifts are usually both inplane and throughplane, but may be thought of as
predominantly one or the other. With inplane shifts it is possible to visualize
on the same image the tissue or fluid which has shifted (or will shift; e.g.,
Figure 7.2), whereas with throughplane shifts the tissue or fluid moving into
the slice comes from adjacent slices. In order to visualize plateaus and border
zones in three dimensions it is therefore necessary to consider adjacent slices
(which are one voxel apart with 3D acquisitions).
