help
489 Subscribers 
help > Denoised Data files
Showing 1-5 of 5 posts
Sep 26, 2016 09:09 AM | Sascha Froelich
Denoised Data files
Dear all,
According to what I've read in this forum (e.g. here http://www.nitrc.org/forum/message.php?m...), CONN should create nifti-files of the denoised data. However, in the folder conn*/results/preprocessing, I do not find any nifti-files. The folder conn*/data doesn't contain any nifti-files, either. But I do have mat-files. However, those mat-files do not seem to contain the voxel timecourses (which I'm interested in). The structs of these mat-files look like this when I load them into MATLAB:
fname: '/Path_to_conn_project/results/preprocessing/cachetmp_DATA_Subject001_Condition001.mat'
issurface: 0
matdim: [1x1 struct]
size: [1x1 struct]
voxels: [262958x1 double]
GM: [ ]
gm: [ ]
scale: 1
voxelsinv: [91x109x91 double]
overwritesoftlink: 1
conditionsweights: {[300x1 double] [300x1 double] [300x1 double] [300x1 double] [300x1 double]}
crop: 0
softlink: 'cachetmp_DATA_Subject001_Condition000.matc'
What can I do to obtain the denoised data in nifti-files? Or alternatively as a mat-file containing the complete voxel timecourses.
Best,
Sascha
According to what I've read in this forum (e.g. here http://www.nitrc.org/forum/message.php?m...), CONN should create nifti-files of the denoised data. However, in the folder conn*/results/preprocessing, I do not find any nifti-files. The folder conn*/data doesn't contain any nifti-files, either. But I do have mat-files. However, those mat-files do not seem to contain the voxel timecourses (which I'm interested in). The structs of these mat-files look like this when I load them into MATLAB:
fname: '/Path_to_conn_project/results/preprocessing/cachetmp_DATA_Subject001_Condition001.mat'
issurface: 0
matdim: [1x1 struct]
size: [1x1 struct]
voxels: [262958x1 double]
GM: [ ]
gm: [ ]
scale: 1
voxelsinv: [91x109x91 double]
overwritesoftlink: 1
conditionsweights: {[300x1 double] [300x1 double] [300x1 double] [300x1 double] [300x1 double]}
crop: 0
softlink: 'cachetmp_DATA_Subject001_Condition000.matc'
What can I do to obtain the denoised data in nifti-files? Or alternatively as a mat-file containing the complete voxel timecourses.
Best,
Sascha
Sep 26, 2016 11:09 AM | Pravesh Parekh - National Institute of Mental Health and Neurosciences
RE: Denoised Data files
Hi Sascha,
If you are looking for the denoised nifti file, go to "Options" in the "Setup" tab and select "Create confound-corrected time-series", The niftiDATA_Subjectxxx_Condition_xxx.nii file would then be in the results/preprocessing folder.
Is that what you are looking for?
Best
Pravesh
Originally posted by Sascha Froelich:
If you are looking for the denoised nifti file, go to "Options" in the "Setup" tab and select "Create confound-corrected time-series", The niftiDATA_Subjectxxx_Condition_xxx.nii file would then be in the results/preprocessing folder.
Is that what you are looking for?
Best
Pravesh
Originally posted by Sascha Froelich:
Dear all,
According to what I've read in this forum (e.g. here http://www.nitrc.org/forum/message.php?m...), CONN should create nifti-files of the denoised data. However, in the folder conn*/results/preprocessing, I do not find any nifti-files. The folder conn*/data doesn't contain any nifti-files, either. But I do have mat-files. However, those mat-files do not seem to contain the voxel timecourses (which I'm interested in). The structs of these mat-files look like this when I load them into MATLAB:
fname: '/Path_to_conn_project/results/preprocessing/cachetmp_DATA_Subject001_Condition001.mat'
issurface: 0
matdim: [1x1 struct]
size: [1x1 struct]
voxels: [262958x1 double]
GM: [ ]
gm: [ ]
scale: 1
voxelsinv: [91x109x91 double]
overwritesoftlink: 1
conditionsweights: {[300x1 double] [300x1 double] [300x1 double] [300x1 double] [300x1 double]}
crop: 0
softlink: 'cachetmp_DATA_Subject001_Condition000.matc'
What can I do to obtain the denoised data in nifti-files? Or alternatively as a mat-file containing the complete voxel timecourses.
Best,
Sascha
According to what I've read in this forum (e.g. here http://www.nitrc.org/forum/message.php?m...), CONN should create nifti-files of the denoised data. However, in the folder conn*/results/preprocessing, I do not find any nifti-files. The folder conn*/data doesn't contain any nifti-files, either. But I do have mat-files. However, those mat-files do not seem to contain the voxel timecourses (which I'm interested in). The structs of these mat-files look like this when I load them into MATLAB:
fname: '/Path_to_conn_project/results/preprocessing/cachetmp_DATA_Subject001_Condition001.mat'
issurface: 0
matdim: [1x1 struct]
size: [1x1 struct]
voxels: [262958x1 double]
GM: [ ]
gm: [ ]
scale: 1
voxelsinv: [91x109x91 double]
overwritesoftlink: 1
conditionsweights: {[300x1 double] [300x1 double] [300x1 double] [300x1 double] [300x1 double]}
crop: 0
softlink: 'cachetmp_DATA_Subject001_Condition000.matc'
What can I do to obtain the denoised data in nifti-files? Or alternatively as a mat-file containing the complete voxel timecourses.
Best,
Sascha
Sep 26, 2016 01:09 PM | Sascha Froelich
RE: Denoised Data files
Dear Pravesh,
thanks, this is basically what I am looking for. I went back to the "Setup" menu and checked the box "Create confound-corrected time-series" and hit "Done". Then I selected "Do not overwrite" as I don't think it is necessary to redo the whole preprocessing procedure (it takes about a day). Afterwards, however, I did not find any new files in the folder results/preprocessing.
Do I have to select "overwrite existing"? If not, what else could be the problem?
Best,
Sascha
thanks, this is basically what I am looking for. I went back to the "Setup" menu and checked the box "Create confound-corrected time-series" and hit "Done". Then I selected "Do not overwrite" as I don't think it is necessary to redo the whole preprocessing procedure (it takes about a day). Afterwards, however, I did not find any new files in the folder results/preprocessing.
Do I have to select "overwrite existing"? If not, what else could be the problem?
Best,
Sascha
Sep 26, 2016 01:09 PM | Pravesh Parekh - National Institute of Mental Health and Neurosciences
RE: Denoised Data files
Hi Sascha,
I think you might have to overwrite the data (you need not do the preprpocessing again). Just click done, allow Conn to import the data, and then re-run the denoising step after overwriting files as necessary. I think that if you do not select the overwrite option, Conn would skip all processing and would not create the files.
Best
Pravesh
I think you might have to overwrite the data (you need not do the preprpocessing again). Just click done, allow Conn to import the data, and then re-run the denoising step after overwriting files as necessary. I think that if you do not select the overwrite option, Conn would skip all processing and would not create the files.
Best
Pravesh
Sep 26, 2016 03:09 PM | Sascha Froelich
RE: Denoised Data files
That was it! Thanks a lot!
Cheers,
Sascha
Cheers,
Sascha