Dear SPMers,
we perform a multiple regression model to clean an fMRI-EPI time series
(wa*nii) from diverse noise effects, we would like to extract from
this a 'cleaned' time series.
The model contains 18 regressors, the 19th is the constant (basically,
motion terms, and certain nuisance ROI time courses).
I added up the residual images (ResI*nii) to the beta_0019.nii.
The resulting images have a mean level (round 120) that differs from the
original
values (round 800), and I could not identify a scaling factor etc. that
would explain this (technically).
Is there another process/step when entring wa*nii images into a standard
multiple regression
model, that changes the 'niveau' of the values? Do the scales of the
ResI*nii and
the constant fit together?
Sorry for this crude stats question,
Philipp
Max Planck Institute of Psychiatry
NMR Research Group
Kraepelinstr. 2-10
80804 Munich
Mail: [log in to unmask]
Phone: 0049-89-30622-413
|