[open] fmri timing

cumhurtas

Hello

We have used opensesame in several EEG experiments and obtained very nice results. Last month I decided to use opensesame with fMRI experiments. I know that this topic has been discussed several times, but I believe that the timing issues such as using the absolute timing with fMRI may still need a bit more clearance.

We are using a fORP interface in order to sync the scanner with the experiment.
In summary, in every full scan (TR), the scanner sends a key response ("t") to the computer and this response triggers the beginning of the experiment. Up until that moment, everything was straightforward with opensesame.

For the experiment loop, I did not want to use a trigger for the beginning of each trial and hence used the absolute timing code that Sebastian has previously suggested.

The trial duration was set to 2700ms and everything seemed to be fine. But while scanning, although at first the scans appeared to be well synchronized with the stimuli which was an easy block design (5 actual stim (5x 2.7 sn) + REST (5 cross) + (5 control stim (5x2.7 sn) ), after a while the sync disappeared and the scanning session ended before the onset of the last three control stim.

I have also tried different backends but nothing changed.

I have realized that psychopy has a code for fMRI experiments (see below), Is it possible to use such code for opensesame or do you have any other suggestions ?

Psychopy fMRI code:

https://gist.github.com/anonymous/162361595551735595ce

my experiment

https://gist.github.com/anonymous/162361595551735595ce

sebastiaan

You can add a short inline script to the start of your trial sequence to ensure non-slip timing:

• http://forum.cogsci.nl/discussion/298

Basically, this script waits at the start of each trial until a fixed period since the start of the experiment has elapsed. The only scenario that I can think of in which this fails (aside from clock drift issues) is when you specify a desired trial duration that is shorter than the actual trial duration: The desired duration should always be slightly longer than the actual duration, so that the script can pad all trials to the same length.

Does this make sense? Have you tried this script, and if so, in what sense did it not work as expected? Do you see the temporal drift also in the OpenSesame log files?

Cheers!
Sebastiaan