Hello SPM experts
I'm wondering multi task series analysis.
Perhaps, this is a low grade question, but I'm confusing.
Can you help me?
I got a data consisting 60scans 6blocks=RARBRC, each block has 10scans.
(R;rest A;task A B;taskB C;taskC)
I made design matrix as follow.
For task A, SOA is 60, and the beginning of scan was 11, duration is
10scans.
For task B, SOA is 60, and the beginning of scan was 31, duration is
10scans.
For task B, SOA is 60, and the beginning of scan was 51, duration is
10scans.
Involving HRF
Analyze was done by the method of GML(regressor was Box-car model)
T contrast was established.
A; 1 0 0 0
B; 0 1 0 0
C; 0 0 1 0
What is the meaning of contrast of B?
For example, voxel N has significantly higher than threshold about
contrast B.
I wonder that I can understand voxel N is activated only(!) in task B.
For the reason I'm wondering, GLM have the least square comparison for
all data.
without exception of scans of task A and C.
Best wishes,
Takefumi Ueno
Psychiatry dep. Kurume Univ. Japan
|