The command line looks ok. The low efficiency is coming from the low number of events per event type.
On 9/27/16 9:52 AM, Alexandre Obert wrote:
Dear all,
I would like to optimize my design by inserting null-events between my trials (in order to better catch the HRF shape with reasonable duration of the experiment) using optseq. I'm not very familiar with this useful tool and I wonder if you could check my setup?
Here my design : I've got 12 trials per run (2 runs per participants). Each of my trial is 34s long and is composed as follow : -An instruction reminder screen of 4s -A video stimulus of 12s to 20s. This screen will be my "stimulus of interest". It's followed by a fixation cross in such a way that their cumulated durations will not exceed 20s -A question about the video followed by a fixation cross in such a way that their cumulated durations will not exceed 20s. -A jitter (a fixation cross), determined by optseq
It's a 2 (A1; A2) x2 (B1; B2) design and I'm interested in these following contrasts: A1 vs A2 B1 vs B2 (The interaction contrast will have an exploratory value in my analysis.)
In order to determine the jitters and the order of the stimuli, I set the optseq as follow:
--ntp 210\ --tr 2\ --psdwin 0 40 1\ --ev evA1B1 34 3\ --ev evA1B2 34 3\ --ev evA2B1 34 3\ --ev evA2B2 34 3\ --ecv 1 1 -1 -1\ --evc 1 -1 1 -1\ --nsearch 10000\ --nkeep 3
I'm not very confident with my setup and if it's correct? I also wonder if I correctly set the ev durations and the psdwin. The stimulus of interest is the video but I set the ev using the trial duration in order to obtain ITI jitters and psdwin to the trial duration. This results in a poor efficiency and I wonder if it's possible to better optimized this design...
Any help will be appreciated...
Regards,
Alexandre
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer