Dear All,
Our fMRI study has used an event-related design to present visual scenes (two regressors with onset times for each of the two image categories). However, we have not modeled any specific events as baseline events, which means there are no onset times that are zero for any of the images.
We are interested in removing the baseline activations to all pictures. I have used the following process to extract the implicit baseline from the individual regressors:
1. Loaded the SPM design after having run the first level analysis
load('SPM.mat', 'SPM')
2. Save the regressor information within the SPM.mat file to another variable
desmat = SPM.xX.X
3. Extract the HRF convolved values for each image for each individual regressor
desmat (:,1) and so on...
Now, can the zero values linked to the images for each regressor be treated as a baseline event for that regressor? If yes, how do I create contrast images which will give me neural activations to all pictures (all non-zero events minus the baseline zero events across the two regressors)?
Any help/suggestions much appreciated.
Best,
Shanker
|