Dear Soroor,
the edge effects are an effect of filtering and whitening. Since both
data and design are treated this way, these effects as well as the non-
zero mean should disappear once you adjust for the constant column.
It is difficult to guess the error without knowing the exact error
message. It doesn't seem to be a mask issue, since extracting data
without adjustment seems to work.
You could try to define a simple F-contrast with a weight of "1" over
your only column and use this contrast to adjust. Technically this
should be the same as adjusting for "everything".
Volkmar
NB: Please direct your answers to the list. Others might also have
useful suggestions or may profit from this conversation.
Am Dienstag, den 11.02.2020, 21:03 +0000 schrieb Soroor Shafieizadegan:
> Dear Volkmar,
>
> thank you so much for useful comment,
> if you meant about model regard to the specification and estimate
> part and design matrix, yes I did it before extracting time series
> but I didn't define any contrast and regressor, so SPM automatically
> considers a column of ones for design matrix. in SPM manual with this
> procedure for extracting the WM time series has considered
> adjustment NaN but the difference my procedure with the manual is in
> using a mask instead of defining coordinate(chapter 38-page:354). and
> when I didn't consider any adjustment (zero) time series is like
> something below, The time-series signal seems to show edge-artifacts
> and have a DC value. that seems not good!
> another question I have, you know how can I extract the mask of
> a special region or all regions from SPM atlas?
>
> Best regards,
> Soroor
|