First of all, the number of stimuli should not influence the estimate of the % change, only its variance (ie, it won't bias you to find a change that is not there). Having said that, you can get some non-linearities if stimuli are too close to each other, so you'll want to make sure to check that the average interstimulus time is about the same. I don't have a strong feeling about whether VRF or efficiency is better -- I don't think I would stress to much over the difference.
doug
raij@nmr.mgh.harvard.edu wrote:
Hi Doug and other Freesurfers,
We are trying to balance the power of different event-related fMRI runs (each consisting of a different experimental condition with different interstimulus intervals) using the optseq2 definition for efficiency. Our outcome metric is the BOLD % signal change amplitude - we expect different runs to show different amplitudes. We use a FIR model. Because of the variable under investigation (interstimulus interval) we must have different number of stimuli in each run.
We are uncertain as to which optseq2 efficiency measure more accurately suits this situation. I know you typically prefer "VRF" over "Efficiency". However, is VRF better suited for reflecting the power of statistical contrast (for example for t-test and F-values) as opposed to "raw" BOLD % signal change? In other words, would "Efficiency" more accurately reflect the balance for just the BOLD % signal change?
The ultimate goal is to make sure that changes in BOLD % signal change amplitude are not due to the different number of stimuli across runs.
Thanks!
Best regards,
Tommi
Tommi Raij, M.D., Ph.D. MGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging Bldg 149, 13th St Charlestown, MA 02129 U.S.A. _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer