Dear all:
I'm a beginner using SPM and resting fMRI. I tried to preprocessed my data using DPARSF, with normalization by using T1 image unified segmentation. My data preparation and directory were arranged according to the online-course. However, DPARSF crashed immediately after finishing realignment with the following error message.
Done 'Realign: Estimate & Reslice'
Done
??? Error using ==> copyfile
No matching files were found.
Error in ==> DPARSF_run at 310
copyfile('co*',['..',filesep,'..',filesep,'T1ImgSegment',filesep,AutoDataProcessParameter.SubjectID{i}])
Error in ==> DPARSF>pushbuttonRun_Callback at 841
[Error]=DPARSF_run(handles.Cfg);
Error in ==> gui_mainfcn at 75
feval(varargin{:});
Error in ==> DPARSF at 33
gui_mainfcn(gui_State, varargin{:});
??? Error using ==> DPARSF('pushbuttonRun_Callback',gcbo,[],guidata(gcbo))
Error using ==> copyfile
No matching files were found.
??? Error while evaluating uicontrol Callback
It seems that some files were missing and problems were happened during T1 segmentation, because DPARSF could work if choosing "Normalize by EPI templates". I apologize that I'm not familiar with Matlab and could not realize what has happened, but I would greatly appreciate all help or suggestions.
Thanks in advance for help regarding this matter.
/carlos
Submitted by YAN Chao-Gan on Fri, 11/27/2009 - 19:08 Permalink
Re
Hi!
If you start with T1 image files instead of T1 DICOM raw files, the prefix of T1 images should be 'co', i.e. T1 image files named in 'co*.img'.
You can input your T1 DICOM images to DPARSF, or rename the files by yourself.
Also, I will added a new utility into DPARSF in several days, which could change the prefix of images easily. You can wait for it.
Best wishes!
Submitted by carloscat on Sat, 11/28/2009 - 01:12 Permalink
Thanks for your answer! But
Thanks for your answer!
But I also arranged DICOM files into the directory "D1Raw" as the online course mentioned, and then chose the function "T1 DICOM to NIFTI".
The T1 images (NIFTI) was then successfully converted by DPARSF into a new directory, "T1Img".
But the error messages remained.
I've checked the T1 images converted by DPARSF, but it seems that these files were not named with the prefix 'co'.
Is there something I've missing?
Submitted by YAN Chao-Gan on Sat, 11/28/2009 - 11:42 Permalink
Re
Hi!
You mean it outputed the .hdr or .img files but without the ones with 'co' prefix?
What is your operation system? Is it Windows Vista?
The function 'T1 DICOM to NIfTI' has been tested in Windows XP and Linux. I have not tested it in Windows Vista.
I will check it in Windows Vista. I am sorry you might just use it in Windows XP or Linux currently.
Best wishes!
Chao-Gan
Submitted by carloscat on Sun, 11/29/2009 - 00:30 Permalink
Thanks for your effort
Thanks for your effort again!
Yes, it is really strange that my outputed .hdr/.img by DPARSF do not have 'co' prefix.
My original DICOM images were aquired by Siemens 3T system and had been outputed by the MRI workstation with following filename "Subject_name.MR.username.study_time.IMA" (for ex. :SU_WAN_TING.MR.HEAD_IBRU.12.1.2009.03.02.20.35.48.765625.39591074.IMA).
And DPARSF would convert it as "Study_date_time_sequence_subject name.img" (for ex. :20090302_191057mpragegrappa2xsuwantings012a1001.img)
But my operation system is Windows XP with Matlab 2008a and SPM8.
I apologize for that I lack the ability for troubleshooting with the source code.
I appreciate your effort and kindness for helping me again.
/Carlos
Submitted by YAN Chao-Gan on Wed, 12/02/2009 - 01:03 Permalink
Re
Hi!
I am sorry I do not know why the 'co*.img' is not created based on your DICOM data.
However, you can use a new Utility in DPARSF to change the prefix of your images. Just click Utilities in DPARSF, and you will find the easy tool to add prefix 'co' to your data.
Best wishes!