| |
Flat Maps
19
The functional flat maps are used by mrFindBorders to identify retinotopic visual areas, i.e.,
only two-dimensional images have to be registered. High-dimensional volumetric warps with
more degrees of freedom are not necessary. Therefore, one reason for using flat maps is the
easier computation.
An important question to ask is if there is a disadvantage when using flat maps. As ex-
plained in chapter 2, the cortex is a folded sheet in three-space. It consists of several layers.
The number of layers varies depending on the location. For example, area V1 can be subdi-
vided into 6 layers and is about 3-4 mm thick (see section 2.3). However, a two-dimensional
functional image that integrates the cortical activity across the cortical layers is no disadvan-
tage since the inplane resolution is about 2 mm x 2 mm, where inplanes are about 3 mm
apart2. Therefore, the resolution is too low to measure the functional signals from different
cortical layers, and the activity is therefore integrated across the layers even when the gray
matter layers are represented in three-space.
3.7 Creating Flat Maps
This section describes how flattened representations of the cortical surface are generated. This
includes the MRI scanning procedures, which produce anatomical data sets, fMRI scanning
for functional data, segmentation and classification of the volumetric anatomy data set, and
finally the actual flattening procedure. In addition, it is briefly explained, how a low-
resolution functional data set is aligned with a high-resolution anatomical data set.
3.7.1 High-Resolution Anatomy Scan
The first anatomical scan is a high-resolution T1-weighted MRI scan. Within about seven
minutes, 124 sagittal slices from the whole brain are measured. Each slice has 256 pixels x
256 pixels. This scan leads to a 3-dimensional volume anatomical data set and is only done
2
The values are given for a 3 T scanner; the inplane resolution for a 1.5 T scanner is about 2.8 mm x 2.8 mm.
|  |
|
| |
|
|
