you should pay attention to threshold and implicit masking options in
your batch. If you have either of them set, SPM may mask out all of your
voxels because they are below threshold or contain zero values.
On Tue, 18 Sep 2007, aaron qin wrote:
> Dear everyone,
> I am processing the resting state data (obtained from GE scanner) with SPM5
> under the
> environment of MATLAB 7.1.
> During the preprocessing procedure, after normalization, bandpass
> filtering (0.01-0.1Hz) was implemented, compared with the former data(after
> the filtered data were centralized (contain positive and negative values).
> After finishing
> the smooth, began to estimate, but pop up the tab:
> ' No inmask voxels ? empty analysis!'
> Then, we add a large enough positive value (eg. 50000) to the data value
> (normalized data, within the mask)which we are interested during the
> bandpass filtering procedure.
> This time the estimation succeed.
> And we discovered that the data type of the header file before and after
> filtering is
> 'INT16-LE'. Data only contains positive values before filtering. After
> filtering, data centralized and contain positive and negative values.
> So, we suppose whether this change will influence the generated data and the
> estimation? And how to figure out the problem?
> Expecting your answer! Thank you very much!
Department of Neurology [log in to unmask]
Universitaetsklinikum Freiburg Phone 49(0)761-270-5331
Breisacher Str. 64 Fax 49(0)761-270-5416
79106 Freiburg http://fbi.uniklinik-freiburg.de/