Hi Maryam,
not sure if Doug is still reading email, but he would be the one to ask. I think you should be able to use mri_vol2surf to sample a volume timeseries into an NxT (nvertices x timepoints) "volume" in .mgz format. You could then read the surfaces into matlab, and the indices into the surface would give you x,y,z and the indices into the NxT "volume" would give you the fMRI values.
cheers, Bruce
On Tue, 20 Nov 2007, Maryam Vaziri wrote:
Hi Cameron,
I want to run a classification analysis on my functional data from all of the cortex using matlab. It is not something complicated but the very fist step is to run a PCA on all the data from all the voxels in the cortex. I am trying to reduce my data set as much as I can as the computation will b every extensive if I have huge files and Hardware-wise impossible to do. So vertex based data the way it is in the surfaces is very big for my purpose. I need volume data. I read my functional volumes to matlab using MRIread and reading the bhdr files in my BOLD directory (I can also read it using fmri_ldbvolume). But the problem with these files is that It is still too big. For reducing it I have to get rid of everything but the cortex. The problem is that I want the whole cortex data, and thus I cannot read just a label. I also need volume data so I cannot read a surface. Moreover I need to reduce my data set so I cannot read the whole brain volume. I also need all the time points so I cannot read the w files from a contrast analysis. What I finaly want is a N*T*4 matrix, containing x,y,z coordinates and values of all (N) voxels in the cortex, in all(T) time points.
I figured that if I can somehow relate the voxels in the ribbon.mgz to the functional volumes I will be all set. But is this impossible? What does the contents of register.dat mean? Is there anyway that I can relate the two files using register.dat? Or maybe somebody has a totally different suggestion?
Bests, Maryam.
On Nov 20, 2007, at 10:32 AM, Cameron Ellis wrote:
Hi Maryam, What type of analysis are you trying to do? It sounds like you are trying to do something unique and not simply follow the fs-fast surface based analysis. Were you able to read the entire functional volume into matlab? It will be hard to limit this matrix to "active" voxels (not 0) because it is unlikely that any voxel in your functional will have a value of 0. I think you should be able to restrict the functional volume to the cortex but I am not sure how to do this. The register.dat file contains the matrix used to relate the anatomic volume to the functional volume but I am also unsure if how to apply this to the ribbon.mgz or the raw functional data and I am not sure that this is the best way to obtain the information you want. I wish I could be more help. I am just not sure of how that data is organized when it is read into matlab (according to scanner coordinates/tal coordinates etc.) Are you sure there is not a way to do what you would like to do without reading the data into matlab? Cameron
Hi again,
I found a file named ?h.ribbon.mgz among the structural files which contains the cortical part of the volume that I want. But I do not know how to relate this structural volume to my functional volumes to get only cortical voxels from my functional volumes. I think that I should use the file register.dat, but I do not know what does the information inside the register.dat mean. What I basically want to do is to read voxels in my functional data, which their corresponding voxel in the ribbon.mgz is not zero and get rid of the voxels that their corresponding voxels in ribbon,mgz is zero.
Bests, Maryam.
On Nov 16, 2007, at 12:28 AM, Cameron Ellis wrote:
Hi Maryam, do you know about the matlab toolbox for freesurfer? it is located at: $FREESURFER_HOME/matlab. all of these scripts are intended to read fmri data into matlab. I think the script you are looking for is MRIread.m or maybe load_analyze.m, load_cor.m or load_dicom.m I think you may be able to load a surface volume using cortical_labeling_tabel.m I actually haven't used these so i could be completely off the mark that these would be useful. Hope this helps. Cameron
Hi,
I am a newbie in using freesurfer so sorry if this is a very trivial question. I want to load my fMRI data into matlab. I want to make a matrix containing coordinates and values of all the voxels in the brain at all timepoints. I also only want the voxels that are in the gray matter and not the voxels that are in the white matter or outside the brain. Where can I get this data and how can I load it to matlab? I have already managed to import all the voxels using MRI- ldbvolume, but this includes white matter and skull. How can I get rid of the rest of the brain and just get the surface?
Bests, Maryam.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer