help
help > RE: Skip denoising
Feb 10, 2015 01:02 AM | Gustavo Pamplona
RE: Skip denoising
Hi Alfonso,
As my data are in subject-space I selected 'Same as functionals' in Setup.Options.AnalysisSpace and 'Implicit mask' in Setup.Options.AnalysisMask. Actually, this is what I was doing before I ask you. However, it doesn't fix the problem.
Because you said that my procedure for preprocessing seems alright, I think I'll have to give you further details so maybe we can fix it.
All the structural and functional images and the GM, WM, and CSF mask images are coregistered and have dimensions 80x80x32. Two ROIs are used in my study, which have dimensions 91x109x91, MNI-normalized, I even thought that this was causing the problem, but it still happens even when I use these ROIs in dimensions 80x80x32 and coregistered.
Do you have any other guess about what can be happening?
Thank you very much!
Best,
Gustavo
Originally posted by Alfonso Nieto-Castanon:
As my data are in subject-space I selected 'Same as functionals' in Setup.Options.AnalysisSpace and 'Implicit mask' in Setup.Options.AnalysisMask. Actually, this is what I was doing before I ask you. However, it doesn't fix the problem.
Because you said that my procedure for preprocessing seems alright, I think I'll have to give you further details so maybe we can fix it.
All the structural and functional images and the GM, WM, and CSF mask images are coregistered and have dimensions 80x80x32. Two ROIs are used in my study, which have dimensions 91x109x91, MNI-normalized, I even thought that this was causing the problem, but it still happens even when I use these ROIs in dimensions 80x80x32 and coregistered.
Do you have any other guess about what can be happening?
Thank you very much!
Best,
Gustavo
Originally posted by Alfonso Nieto-Castanon:
Dear
Gustavo,
This error indicates that your functional data (for the first subject/session) contains no voxels of interest, so the most likely reason would be perhaps a misalignment between the mask used for functional analyses (defined in Setup.Options.AnalysisSpace and Setup.Options.AnalysisMask) and your functional data (entered in Setup.Functionals). If your data is in subject-space (not MNI-normalized) then simply select 'Same as functionals' in Setup.Options.AnalysisSpace and 'Implicit mask' in Setup.Options.AnalysisMask. Alternatively if your data is normalized please check whether the coregistration between the chosen analysis mask and your functional data seems correct (e.g. using SPM 'check coregistration') and select an alternative mask otherwise.
Let me know if that seems to fix this issue (the 'skipping the denoising step' procedure you are using looks perfectly correct, no issues there)
Best
Alfonso
Originally posted by Gustavo Pamplona:
This error indicates that your functional data (for the first subject/session) contains no voxels of interest, so the most likely reason would be perhaps a misalignment between the mask used for functional analyses (defined in Setup.Options.AnalysisSpace and Setup.Options.AnalysisMask) and your functional data (entered in Setup.Functionals). If your data is in subject-space (not MNI-normalized) then simply select 'Same as functionals' in Setup.Options.AnalysisSpace and 'Implicit mask' in Setup.Options.AnalysisMask. Alternatively if your data is normalized please check whether the coregistration between the chosen analysis mask and your functional data seems correct (e.g. using SPM 'check coregistration') and select an alternative mask otherwise.
Let me know if that seems to fix this issue (the 'skipping the denoising step' procedure you are using looks perfectly correct, no issues there)
Best
Alfonso
Originally posted by Gustavo Pamplona:
Dear CONN experts,
I want to analyze some almost raw data, so I only performed the realignment, coregister, and spatial smoothing in SPM before insert the data in CONN fields. It may seem strange but I don't want to perform the denoising to get the access to the first-level analysis, so I did not put any confounds to improve the original data, chose no detrending and wrote [0 Inf] in band-pass filter. I know I'm gonna get very noise data, but that's intentional.
However, in the 3rd step of denoising, I got this error message:
ERROR DESCRIPTION:
Error using conn_process (line 877)
Empty data (samples=400,voxels=0) in file C:\Gustavo\Imagens\Neurofeedback\dados_angelo2\data\DATA_Subject001_Session001.mat
Error in conn_process (line 17)
case {'preprocessing_gui','denoising_gui'}, disp(['CONN: RUNNING PREPROCESSING STEP']); conn_process([1.5,2,6:9],varargin{:});
Error in conn (line 2181)
conn_process('preprocessing_gui');
Error in conn_menumanager (line 102)
feval(CONN_MM.MENU{n0}.callback{n1}{1},CONN_MM.MENU{n0}.callback{n1}{2:end});
CONN v.14.n
SPM8 + Beamforming DEM FieldMap MEEGtools
Matlab v.2013a
Could someone please show me what's wrong and tell me how to skip the denoising step?
Thanks in advance!
I want to analyze some almost raw data, so I only performed the realignment, coregister, and spatial smoothing in SPM before insert the data in CONN fields. It may seem strange but I don't want to perform the denoising to get the access to the first-level analysis, so I did not put any confounds to improve the original data, chose no detrending and wrote [0 Inf] in band-pass filter. I know I'm gonna get very noise data, but that's intentional.
However, in the 3rd step of denoising, I got this error message:
ERROR DESCRIPTION:
Error using conn_process (line 877)
Empty data (samples=400,voxels=0) in file C:\Gustavo\Imagens\Neurofeedback\dados_angelo2\data\DATA_Subject001_Session001.mat
Error in conn_process (line 17)
case {'preprocessing_gui','denoising_gui'}, disp(['CONN: RUNNING PREPROCESSING STEP']); conn_process([1.5,2,6:9],varargin{:});
Error in conn (line 2181)
conn_process('preprocessing_gui');
Error in conn_menumanager (line 102)
feval(CONN_MM.MENU{n0}.callback{n1}{1},CONN_MM.MENU{n0}.callback{n1}{2:end});
CONN v.14.n
SPM8 + Beamforming DEM FieldMap MEEGtools
Matlab v.2013a
Could someone please show me what's wrong and tell me how to skip the denoising step?
Thanks in advance!
Threaded View
| Title | Author | Date |
|---|---|---|
| Gustavo Pamplona | Feb 4, 2015 | |
| Alfonso Nieto-Castanon | Feb 8, 2015 | |
| Gustavo Pamplona | Feb 10, 2015 | |
| Alfonso Nieto-Castanon | Feb 11, 2015 | |
| Gustavo Pamplona | Feb 10, 2015 | |
| Gustavo Pamplona | Feb 10, 2015 | |