About critical timing on fMRI experiment
Julian Gaviria
Hello,
I am performing an fMRI experiment with a flanker task which is comprised by one block with 80 trials (congruent and incongruent conditions) with a duration of 1s per trial. And there is a fixation object between trials with different duration (jittering the fixation objects).
Questions
1. After performing my experiment, I read (http://imaging.mrc-cbu.cam.ac.uk/imaging/EprimeTiming) that for an fMRI experiments it is "necessary" to use a pre-release value (100 ms) from the ‘Duration/Input’ input tab of the properties of an object. Is it really necessary? even if I have long fixations among the trials (stimuli)?
2.After reading the user guide and some chats from the google group, I've got the impression that if I need to know as accurate as possible the time during the stimuli are presented and the fixation times, I should log the "Starttime" and "Finishtime" options of both of them. Is it right? In case I did not log them. Is there any chance to retrieve the information? (I run my fMRI experiment with the standard options)
It would be nice a clearer suggestion about it, since I can not interprete the info form the data I acquired (I only have the standard into from log ).
P.D. For the behavioral analysis, there is no problem (the reaction time and the accuracy are fine). But for the fMRI analysis I do need to know exactly the time length for each trial because I need to create a model.
Thank you in advance
David McFarlane
timing performance of your program. If you do use PreRelease, you might
as well use as much as possible (I go over this in my online course).
In fact, starting with EP2.0.10, the default value of PreRelease is
"(same as duration)".
2) Please see https://groups.google.com/d/topic/e-prime/OeiZ00V9SRc for
a discussion of time audit measures (StartTime, FinishTime, etc.). No,
there is no good way to retrieve these values if you did not log them
during the run. I find TargetOnsetTime the most useful measure for
evaluating fMRI timing, that is what I can actually control (I go over
this in a lesson in my online course).
That all said, note that you may use two different strategies for
synchronizing your program timing with the scanner: (1) Set all
stimulus objects to use Cumulative timing mode, then synchronize your
E-Prime program once at the start of the scan; (2) Resynchronize your
E-Prime program with a TR-pulse from scanner for each trial.
---------------
David McFarlane
E-Prime training online:
http://psychology.msu.edu/Workshops_Courses/eprime.aspx
Twitter: @EPrimeMaster (https://twitter.com/EPrimeMaster)