| |
Introduction to mrFindBorders
62
Pre and Post Automatic Deformation
This item in the Tools menu contains commands that can be used both before and after the
deformation process.
·
It is possible to get a phase value at any point in the currently active image by clicking
at the point’s location in the image. In addition, the location's coordinates are also dis-
played.
·
The energy can be calculated for the current data set. For the calculation, the weights
of the energy components for the current data set, which are specified in the options
menu, are used. The total energy, as well as the percent contribution of each compo-
nent, is given. The matrix of squared differences is displayed with a color map that
shows low values as dark colors and high values as bright colors. The matrix of
squared differences is calculated based on the weighted sum of two difference matri-
ces. One is calculated by calculating the matrix of squared differences between the
measured phase-wedge image and the corresponding atlas images. The second one is
calculated in the same way but for the ring-stimulus case. The displayed image is an
error map. It is useful to see where the measured images are similar to the correspond-
ing atlas images. In addition to the energy, a quality measure can be calculated. This
measure is not based on the phases in the images, but only on the resulting visual-area
ROIs. Only these ROIs are of interest when the visual areas are to be identified and
therefore only they are to be evaluated. Gold-standard reference visual-area ROIs are
used to calculate this quality measure. More details about energy and quality calcula-
tion can be found in later chapters.
·
Atlas images can be deformed manually by assigning displacement values to control
points. Figure 31 (top row) shows that the user can choose a control points first. The
selected control point is displayed (large +) together with the currently assigned dis-
placement values for the selected control point (small +). The user can specify new
displacement values by clicking at a pixel position. The pixel value at that position
will end up at the current control point after deformation. Finally, the atlas is deformed
based on the new displacement fields. The resulting visual-area ROIs can be seen in
|  |
|
| |
|
|
