| |
Flat Maps
20
once on a certain person. In [Wandell, et al., 2000 #38], the voxel size is 0.9 mm x 0.9 mm x
1.2 mm on a GE 1.5 T Signa scanner. This high-quality anatomical data set (called vAnatomy
= volume anatomy) serves as an anatomical coordinate frame. It is needed because functional
data from each subject are collected in several sessions. By registering the low-resolution data
from each session with the high-resolution full-brain vAnatomy data set, the data from various
sessions can be brought into the same coordinate space [Demb, 1997 #32].
3.7.2 Low-Resolution Inplane Anatomy Scan and Functional
Scan
When functional data is to be recorded during a session, T1-weighted low-resolution inplane
anatomical (structural) images are taken first. Typical values are: images have 128 pixels x
128 pixels; inplane resolution 2 mm x 2 mm; inplanes separated by 3 mm; standard spin-echo
sequence. Typically 23 coronal inplanes are recorded in less than 2 minutes with a clinical
GE 3.0 T Signa scanner; many repetitions are acquired within this time period to improve the
signal-to-noise ratio.
Then, the observer has to fixate on various stimuli and BOLD fMRI signals are acquired.
This yields in 23 coronal planes (128 x 128 pixels). The inplane scans are intrinsically aligned
with the functional planes. This is possible since the subject is supposed to not move during
the session. The fMRI data is recorded using a T2*-sensitive gradient recalled echo pulse se-
quence with a spiral readout. It takes about 3 seconds to record data from the whole set of
slices, i.e., each slice is acquired every 3 seconds [Demb, 1997 #32] (they used a 1.5 T scan-
ner).
3.7.3 Segmenting High-resolution Anatomical Data
The aim of this step is to find the location and contiguity of the gray matter within the high-
resolution anatomical data. [Teo, et al., 1997 #8] describe that white matter and cerebral spi-
|  |
|
| |
|
|
