Dear SPM experts, I am trying to use an open source dataset for analysis on resting brain dynamics.
The dataset is composed by an anatomical image T1w and functional images (210 volumes) for each subject. I need to segment the T1 in order to obtain the Destrieux and Desikan atlas. I can segment the T1 with freesurfer or CAT12, it is not a problem with both of them. However I have no idea how to use the atlases for extracting ROI BOLD time series. My purpose it to obtain a simple matrix N (Region of the atlas) x M (BOLD time point) to do additional analyses.
My analysis Pipeline with SPM12 is the following: I realign the images, reslice in time, coregister, segment, normalize and smoothing.
First of all: how do I regress out motion, cardiac and respiratory artifact so that I know that when I am extracting the bold time series I am extracting a cleaned BOLD signal?
Second: can I read mgz file from Freesurfer in SPM12?
Third: Once I have coregistered both the structural, the functional and the atlas in the same space, how can I use the the Freesurfer atlas to extract the data creating a simple matrix ROI of altas x BOLD samplig points.? In my case I have 210 volumes, and 114 regions therefore I need to obtain a matrix 114 x 210.
I exstensively check online, there are some suggestions of creating the ROI with the sphere with Marsbar, or using pickatlas, but I must say that those examples do not help at all in my case.
I believe that a lot of SPM users had my same problem, therefore I believe that there should be a solution already made for this. Unfortuntaley I cannot use other software like FSL or AFNI since I use window system in this case. I run Freesufer via CBRAIN server.
Therefore I am asking you, is there a way to solve this problem to obtain a data matrix roi x bold sampling points derived from an volumetric atlas from freesurfer with SPM?
Thanks for your help
|