"Index exceeds matrix dimensions"

Submitted by yhcha@mednet.uc... on Fri, 03/18/2011 - 12:51

Hi, I'm fairly new to functional imaging and am very interested in finding a resting state analysis program that is user friendly. I've encountered the following error message, which I can't figure out despite listening to the online tutorial several times. The way that I've entered the data looks just like the sample screens except that we use 36 slices and have 120 volumes of data.
Thank you for any advice on this topic.
Yoon-Hee

?? Index exceeds matrix dimensions.

Error in ==> DPARSF_run at 215
InputFilename=[AutoDataProcessParameter.DataProcessDir,filesep,'FunRaw',filesep,AutoDataProcessParameter.SubjectID{i},filesep,DirDCM(3).name];

Error in ==> DPARSF>pushbuttonRun_Callback at 938
[Error]=DPARSF_run(handles.Cfg);

Error in ==> gui_mainfcn at 96
feval(varargin{:});

Error in ==> DPARSF at 40
gui_mainfcn(gui_State, varargin{:});

??? Error while evaluating uicontrol Callback

Submitted by YAN Chao-Gan on Fri, 03/18/2011 - 14:37

Hi!
It seems that DPARSF can not find the your DICOM files.
Could you check what kind of files do you have under XXXX\FunRaw\Subject01\?
Best wishes!

Submitted by yhcha@mednet.uc... on Sat, 03/19/2011 - 00:45

Thank you for your post. The files that come off of our scanner are plain text files. In this case, there are 120 volumes at 372kb each. For regular fMRI analysis, I do a DICOM to Nifti conversion using dcm2niigui and then use the .nii files for subsequent analysis. These plain text files don't have the ".dcm" suffix from your example. I am doing the analysis on a MacOSX using Matlab 2009b and spm8.
Thank you so much for your help. I really appreciate it.
Yoon-Hee

Submitted by YAN Chao-Gan on Sun, 03/20/2011 - 15:17

Hi!
Firstly, DPARSF works not so well with MacOS, but it works well with Windows and Linux.
Secondly, If you use dcm2niigui to convert your DICOM files, plese choose the output format as ".img/.hdr". Then you can skip the step of "EPI DICOM to NIFTI" in DPARSF.

Best wishes!

Submitted by yhcha@mednet.uc... on Mon, 03/21/2011 - 13:49

Thank you for this insight. I will find a way to create a Windows platform and try this again and will report back to you. Thanks for your great advice.
Yoon-Hee

Submitted by yhcha@mednet.uc... on Mon, 03/28/2011 - 13:10

Hi again,
I was able to install everything onto a Windows platform (spm8 Matlab 7b) and rerun the analysis. Fortunately, I don't get the same error. But, the new error says: The detected time points of subject "Sub001" is: 0, it is different from the predefined time points: 120. Please check you data!
In the FunRaw folder, I have 120 hdr/img pairs. The labels looks like this: 20101118_162154BlankMdDSACs003a001_001.hdr
I also found that when I tried to convert my mprage raw data to hdr/img format, I only get one file, not 192. There are 192 images on my mprage.
Thanks again for your help. It looks like you were right about the original problem. It was probably the Windows issue.
Yoon-Hee

Submitted by dongzy08 on Sat, 04/02/2011 - 09:42

The first problem, make sure the file directory is X:\...\FunRaw\Sub001\..., the hdr/img pairs are under this directory.
The second problem, maybe you convert the raw data to 4D format which contains the time dimension. Which converting tool you use? If it is dcm2niigui, please make sure the option is 3D nifti.