Dear experts,
I am currently trying to analyse event-related data. I read the manual chapter on this (with the faces data), and watched the demo video from SPM website (talk gave by Steeve Fleming).
However, in his talk, Steeve said he used normalized and smoothed files, but it seems he did not use Slice Timing correction (did not see the "a" in the file name), as specified in the manual.
I was then wondering if I should apply slice timing correction to my dataset? I also read about (controversial) slice timing, as well as on using it first or after realignment but did not find a clear response (probably because there is no clear response?).
As explain in both manual and Steeves talk, I may also use time derivative (but not dispersion derivative??) when setting my 1st level. But should I specify slice timing at the same time than time derivative?
Which of the following options would you recommend?
1) NO slice timing + time derivative
2) Slice Timing first + NO derivative
3) Slice Timing first + time derivative
4) Slice Timing after realignement + NO derivative
5) Slice Timing after realignement + time derivative
Acquisition specifications (Philips 3T Achieva): TR= 2000 ms, TE=30 ms, 32 ascending (interleaved?) axial slices.
If I have to use slice timing, what would be the good order to enter my slices? I assume it will be 1 3 5 ... 31 2 4 6...32?
Also, when I will specify my 1st level, is that right to set unit as seconds, microtime at 32 and microtime onset at 16?
Thanks in advance for your advices.
BW,
Yann
|