| |
182
8 Discussion and Future Work
8.1 Overview
The validation section has shown that it is possible to segregate retinotopic visual areas by
using a deformable model. mrFindBorders has produced good results for different data sets.
For difficult data, i.e., flat maps that are very noisy and/or dissimilar to the atlases, the user
can use tools to deal with those problems. For example, he/she can specify points (landmarks)
in the boundaries of the measured phase-wedge image.
Therefore, the main goal of creating a working tool has been achieved. However, no tool
is perfect and improvements are possible. Although it was tried to make the program as effi-
cient and fast as possible, the main goal was to have a running and useful tool at the end. In-
stead of speeding the program up by trying out different optimization algorithms, it was found
to be more helpful to implement some practical features, e.g., the iso-angle and iso-
eccentricity creation and exporting functionalities. These features really help users of mrFind-
Borders to create good and useful results. Iso-angle lines, made by mrFindBorders, are already
utilized by Stanford researchers. They found mrFindBorders to be very practical and iso-angle
lines made by mrFindBorders have produced better results in the study of cortical magnifica-
tion than iso-angle lines that were drawn by hand.
|  |
|
| |
|
|
