Hello,
I am working on midbrain activity analysis, and we are modeling every
subject in their native space. We have extracted BOLD timeseries for each
subject, and the average timeseries across subjects seems very reasonable.
However, our betas do not make sense when compared to these time-series. We
averaged each regressor across all subjects, and the values do not match
the activity profile shown in the timeseries. Additionally, several of the
regressors have negative values.
We have also attempted using the substantia nigra as an explicit mask
during estimations (this is the ROI of the present study). This was done
because we were worried that the large magnitudes of cortical activity
might be affecting the nonsensical betas from midbrain, which have very low
levels of activity. However, when we use this explicit mask, we receive an
error during first levels stating "Error: No Significant Voxels." I am not
quite sure what this error means and am having difficulty resolving any of
the aforementioned problems. Any insights would be greatly appreciated.
Thank You,
Anthony
|