 |
 |
HI, I'm trying to analyse sparse sampling data involving multiple volumes per effective TR (e.g., ISSS). For each effective TR of 12.5s, there are 5 TRs of 2.5s, consisting of 2 TRs of silent period and 3 TRs of image acquisition. Stimuli are presented during the silent period. There are 70 repetitions resulting in 210 volumes in total. Since FSL assumes that only one volume is acquired per effective TR, I have the following questions: 1) what should the FEAT TR be? According to (http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FEAT/FAQ) , we should "set the TR to be the actual time from the start of one volume to the start of the next", but in our case, the volumes were clustered together and then followed by a silent period. When I tried 12.5s and 2.5s, the design matrices looked incorrect. I then tried 4.16s (12.5/3 volumes), which generated a correct design matrix, but I'm not sure if this is correct, since it's neither the volume TR or effective TR. 2) For the timing onsets, do I need to be increase them by (TR given to FEAT)/2 - (real TR)/2? Does it make sense to use this formula when the volumes are clustered together? Many thanks, Catherine