Optimizing event design patterns
The distribution of inter-stimuli interval (ISI) durations is a crucial issue in order to optimize event protocols.
In this regard, we advice to consult Rick Henson‘s pages on the efficiency of experimental designs in fMRI.
Matthew Brett wrote a Matlab routine to generate exponential distributions of ISI.
We used his routine to write a function to generate – for N desired runs – an exponential distribution of IIS.
Two annotated .m files are provided :
ISIGeneration.m : The function that generates the ISI distribution.
MakeTiming.m : The code that uses ISIGeneration.m
MakeTiming.m will generate N runs with only the inter-stimuli intervals (correctly randomized). It does not include the stimulus durations. Moreover, the N runs will have the same length in time, expressed in number of MRI-slices.
The code is annotated and can be easily modified in order to set the number of runs, the minimum duration, the average ISI duration, the duration of an MRI-slice, …