Hi Chao, Again, I recommend that you join the HCP-Users email list where topics like this are discussed-- join here: http://humanconnectome.org/contact/#subscribe
1) 91282 are the surface vertices for RH and LH, plus the subcortical voxels. 1200 are the timepoints. The wb_command -file-information command will give you a summary of the mapping in the file. See http://www.mail-archive.com/hcp-users@humanconnectome.org/msg00627.html 2) Yes, I believe the data in rfMRI_REST1_LR_Atlas_hp2000_clean.dtseries.nii can be used straight, but you might want to concatenate across all the runs of the subject before you do further processing, see http://www.mail-archive.com/hcp-users%40humanconnectome.org/msg00446.html for advice on how to do that. 3) wb_view can read label or metric files in GIFTI or CIFTI formats (*.dlabel, *dscalar) that map data on the surface, but could you tell us more about how you are doing your clustering so that we can give you the right commands to create the appropriate files? 4) See 1) Im forwarding the thread to the list so that others there can confirm/add to the discussion. Best, Jenn Jennifer Elam, Ph.D. Outreach Coordinator, Human Connectome Project Washington University School of Medicine Department of Anatomy and Neurobiology, Box 8108 660 South Euclid Avenue St. Louis, MO 63110 314-362-9387 el...@pcg.wustl.edu www.humanconnectome.org From: Chao Liu [mailto:chao....@brunel.ac.uk] Sent: Monday, October 20, 2014 6:12 AM To: Jennifer Elam Subject: Re: [HCP-Info] [Contact Form]: Model free analysis methods for neuroimaging data (e.g. fMRI) Hi Jennifer, Thank you very much for your kind help! All the information you provided are very helpful. And now I can load the dtseries data into Matlab according to your instructions. In oder to make sure what I get was right, I am wondering could you please help point out whether the following facts I got are right? For subject 100307 resting state fMRI data rfMRI_REST1_LR_Atlas_hp2000_clean.dtseries.nii (under folder 100307/MNINonLinear/Results/rfMRI_REST1_LR): 1) I got a 91282 by 1200 matrix. The 91282 are the voxels and the 1200 are the corresponding time series points. 2) The above loaded data can be straightly used without applying further preprocessing (for model free analysis such as clustering). Now the data can be loaded and analysed. The final thing for a complete analysis is the results visualisation. For example, if I use clustering method to group the above 91282 voxels into clusters with members in each cluster having similar BOLD patterns. Then I would like to visualize these clusters (one cluster one color) using wb_view tool. So: 1) Whats the format of input should be given to wb_view so that it can plot the colorful patches on the brain model for the dtseries.nii data? 2) How to get the information (such as brain area name, hemisphere) for each voxel of 91282 using some matlab functions or wb_command rather than using the wb_view GUI? Thank you very much for your time and please feel free to use the contents in my email in your FAQ section if you find them helpful to other researchers. Cheers, Chao On 16 Oct 2014, at 17:32, Jennifer Elam <el...@pcg.wustl.edu> wrote: Hi Chao, First, I recommend that you join the HCP-Users email list where topics like this are discussed-- join <http://humanconnectome.org/contact/#subscribe> here. There is a specific thread with attached matlab functions that you can use for importing CIFTI files (e.g., *.dtseries.nii) into matlab: <http://www.mail-archive.com/hcp-users%40humanconnectome.org/msg00623.html> http://www.mail-archive.com/hcp-users%40humanconnectome.org/msg00623.html Here¹s also some more detailed instructions that do the same thing as the ciftiopen and ciftisave code attached in the above thread: You need a couple of prerequisites: 1) Workbench v1.0 (available <http://humanconnectome.org/software/get-connectome-workbench.html> here) needs to be installed on the system 2) A matlab set of functions called the GIFTI toolbox by Guillaume Flandin ( <http://www.artefact.tk/software/matlab/gifti/> http://www.artefact.tk/software/matlab/gifti/) To convert and load a file into MATLAB, you do the following (in MATLAB to make things convenient) 1) system(['wb_command -cifti-convert -to-gifti-ext ' NAME_OF_FILE ' ' SOME_RANDOM_FILE_NAME '.gii']); 2) GIFTI_FILE = gifti([SOME_RANDOM_FILE_NAME '.gii']); 3) CIFTI = GIFTI_FILE.cdata; 4) delete([SOME_RANDOM_FILE_NAME '.gii'],[SOME_RANDOM_FILE_NAME '.gii.data']); What the steps do: 1) Converts the NIFTI-2 file type to a GIFTI external file. 2) Loads the gifti external filetype from (1) into MATLAB 3) "Read-out" the Cifti timeseries from gifti external "structure" 4) Delete temporary files made to read in. To save out the Cifti timeseries in an easy way: 1) GIFTI_FILE.cdata = CIFTI_modified 2) save(CIFTI_modified,[OUTPUTNAME '.gii'],'ExternalFileBinary'); 3) system(['wb_command -cifti-convert -from-gifti-ext ' OUTPUTNAME '.gii' OUTPUTNAME '.dtseries.nii']); 4) delete([OUTPUTNAME '.gii'],[OUTPUTNAME '.dat']); What the steps do: 1) Load the modified cifti file back into the gifti "structure" for easy saving. 2) Save the gifti as a gifti external binary 3) Use wb_command (in MATLAB) to convert from BLAHBLAH.gii to BLAHBLAH.dtseries.nii 4) Delete the intermediary .gii and .dat files Let me know which of these you find most helpful. I am currently writing this up for a FAQ page and it would help to have your feedback. Best, Jenn Jennifer Elam, Ph.D. Outreach Coordinator, Human Connectome Project Washington University School of Medicine Department of Anatomy and Neurobiology, Box 8108 660 South Euclid Avenue St. Louis, MO 63110 314-362-9387 <mailto:el...@pcg.wustl.edu> el...@pcg.wustl.edu <http://www.humanconnectome.org/> www.humanconnectome.org -----Original Message----- From: <mailto:info-boun...@humanconnectome.org> info-boun...@humanconnectome.org [ <mailto:info-boun...@humanconnectome.org> mailto:info-boun...@humanconnectome.org] On Behalf Of <mailto:i...@humanconnectome.org> i...@humanconnectome.org Sent: Thursday, October 16, 2014 8:33 AM To: <mailto:i...@humanconnectome.org> i...@humanconnectome.org Subject: [HCP-Info] [Contact Form]: Model free analysis methods for neuroimaging data (e.g. fMRI) Name: Chao Liu Email: <mailto:chao....@brunel.ac.uk> chao....@brunel.ac.uk Subject: [Contact Form]: Model free analysis methods for neuroimaging data (e.g. fMRI) Comments: Hi, I feel very lucky to have the change to explore the open accessed HCP data which is a great project. I am wondering if there is a straightforward way to load the 4D data into matlab environment (e.g. compact rfMRI nii file) to let my methods applied. I have tried several tools like SPM and FSL to load those nii file but none of them worked. I would really appreciate any help! Thank you very much. Chao _______________________________________________ Info mailing list <mailto:i...@humanconnectome.org> i...@humanconnectome.org <http://lists.humanconnectome.org/mailman/listinfo/info> http://lists.humanconnectome.org/mailman/listinfo/info _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
