I was fortunate enough to attend your training session last summer. During my time there, I was told by several instructors not to model separately for rest blocks of fMRI data in a block design analysis and look at task-rest activity. I was told this would interfere with the modeling because it will cause variance problems and redundant subtraction due to averaging. I was surprised by this suggestion because I was always taught that you should model rest and subtract. During recent conversations with colleagues, this topic came up again and we were wondering if you could suggest some literature regarding this issue to help our discussion and optimizing analyses with FSL. Thanks for your time.
|