 |
 |
Dear Jim User specified multiple regressors are not convolved with the HRF. Note however that by default all regressors, except for the constant, are considered as being of interest, and are being used to define the mask in which the AR(1) model is estimated. This can be changed, but unfortunately not via the batch interface. Most people don’t care about this but if you decide to do so, you should modify the following fields in SPM.mat before model estimation: I.e. load SPM.mat SPM.xX.iC=ColumnsOfRealInterest SPM.xX.iG=ColumnsOfNoInterest SPM.xX.iB=TheConstantRegressor save SPM.mat SPM Best Torben > On 19 Dec 2016, at 22.06, Ronald Smith <[log in to unmask]> wrote: > > Hi! > > I am using SPM8, and am interested in censoring high motion scans as suggested in > > Siegel, et al. “Statistical Improvements in Functional Magnetic Resonance Imaging Analyses Produced by Censoring High-Motion Data Points.” Human Brain Mapping 35, no. 5 (May 2014): 1981–96. > > I have seen it suggested here on this listserv that an effective way to do this would be to incorporate these as regressors of no interest in the design matrix. I am a little confused, however, as doing so would convolve these with the hrf (correct?), which does not seem ideal, as the changes in signal are not haemodynamic in origin. I would be grateful if someone might clarify, for a SPM novice, whether my concerns are on the mark, and what the best approach would be (other than say using ArtRepair, which I understand interpolates scans rather than drop them). > > Thanks! > > Jim