Dear Rodolphe,
thanks for the quick reply! Yes, that's certainly the best solution, as the "ghost" volumes would also be contained in the following 4D niis generated during preprocessing, e.g. normalised ones (which is no problem as such, but nonetheless suboptimal, say someone else wants to work with the data). Just thought the 4D mean.nii is somewhat useless in the context of realignment ;-)
For those who prefer the GUI, SPM's "3D to 4D File Conversion" also does do the job if you select relevant volumes from a 4D, which are then saved into another 4D.
Best,
Helmut
|