help
help > RE: default_ss matrix with zeros
Mar 30, 2023 08:03 AM | silvia de francesco
RE: default_ss matrix with zeros
Thank you very much for the answer, we are performing only the
first level analysis on single subjects.
We have the FreeSurfer segmentation in native space, we would obtain the Z for the 32 networks of CONN (i.e. networks.DefaultMode.MPFC; networks.DefaultMode.LP; etc.) is there a correspondence between the Freesurfer ROIs and the CONN networks? So I could use the freesurfer ROIs as subject-specific ROIs.
Thank you for your help
Silvia
Originally posted by Alfonso Nieto-Castanon:
We have the FreeSurfer segmentation in native space, we would obtain the Z for the 32 networks of CONN (i.e. networks.DefaultMode.MPFC; networks.DefaultMode.LP; etc.) is there a correspondence between the Freesurfer ROIs and the CONN networks? So I could use the freesurfer ROIs as subject-specific ROIs.
Thank you for your help
Silvia
Originally posted by Alfonso Nieto-Castanon:
Dear
Silvia,
The network ROIs that you seem to be using are defined in MNI-space, so typically you would use them in combination with preprocessing pipelines that bring your functional and anatomical data to MNI-space as well. A subject-space preprocessing pipeline is typically used in combination with subject-specific ROIs (e.g. functional localizers) which are also defined in each subject's space.
In general (with the exception of gray/white/CSF ROIs) the preprocessing pipeline will not touch/modify your ROI files, it will only work on functional and anatomical images, so you need to define ROIs that are in the same space as some version of your functional images, typically the fully preprocessed functionals (in Setup.other you can also specify multiple versions of your functional data, e.g. defined in different spaces, to be used in cases where you need to combine ROIs that are also defined in multiple different spaces).
Hope this helps
Alfonso
Originally posted by silvia de francesco:
The network ROIs that you seem to be using are defined in MNI-space, so typically you would use them in combination with preprocessing pipelines that bring your functional and anatomical data to MNI-space as well. A subject-space preprocessing pipeline is typically used in combination with subject-specific ROIs (e.g. functional localizers) which are also defined in each subject's space.
In general (with the exception of gray/white/CSF ROIs) the preprocessing pipeline will not touch/modify your ROI files, it will only work on functional and anatomical images, so you need to define ROIs that are in the same space as some version of your functional images, typically the fully preprocessed functionals (in Setup.other you can also specify multiple versions of your functional data, e.g. defined in different spaces, to be used in cases where you need to combine ROIs that are also defined in multiple different spaces).
Hope this helps
Alfonso
Originally posted by silvia de francesco:
Dear developer,
I am using CONN 21.a, I am interested the first level ROI-to-ROI analysis in native space analysis. I run my analysis with the defaults_ss option, the output matrix contains only zero values and the registration does not seem well performed but no errors are reported. If I run the analysis in default_MNI everything is ok.
Below my script
conn_file = [DefaultMode.mat'];
CONN_x.filename = conn_file;
CONN_x.Setup.isnew = 1;
CONN_x.Setup.functionals{1}{1}=['RS_fMRI.nii'];
CONN_x.Setup.structurals{1}=['mri/T1.mgz'];
CONN_x.Setup.RT=2.4;
CONN_x.Setup.nsubjects=1;
CONN_x.Setup.voxelresolution=4;
CONN_x.Setup.preprocessing.steps='default_ss';
CONN_x.Setup.preprocessing.sliceorder='interleaved (bottom-up)';
CONN_x.Setup.preprocessing.diffusionsteps=40;
CONN_x.Setup.done = 1;
CONN_x.Denoising.done = 1;
CONN_x.Analysis.done = 1;
CONN_x.Results.done = 1;
CONN_x.Analysis.type = 1;
CONN_x.Analysis.name='DefaultMode';
CONN_x.Analysis.sources = {'networks.DefaultMode'}
CONN_x.QA.plots = {'QA_NORM structural', 'QA_NORM functional', 'QA_NORM rois', 'QA_REG functional', 'QA_REG structural', 'QA_COREG functional', 'QA_TIME functional', 'QA_TIMEART functional', 'QA_DENOISE histogram', 'QA_DENOISE timeseries', 'QA_DENOISE FC-QC', 'QA_DENOISE scatterplot', 'QA_SPM design', 'QA_SPM contrasts' ,'QA_SPM results', 'QA_COV'}
conn_batch(CONN_x);
Where is the error?
I also tried to register the structural and the functional images (they are acquired in the same session) before run CONN but the problem has not been solved.
Thank you for your help!
Silvia
I am using CONN 21.a, I am interested the first level ROI-to-ROI analysis in native space analysis. I run my analysis with the defaults_ss option, the output matrix contains only zero values and the registration does not seem well performed but no errors are reported. If I run the analysis in default_MNI everything is ok.
Below my script
conn_file = [DefaultMode.mat'];
CONN_x.filename = conn_file;
CONN_x.Setup.isnew = 1;
CONN_x.Setup.functionals{1}{1}=['RS_fMRI.nii'];
CONN_x.Setup.structurals{1}=['mri/T1.mgz'];
CONN_x.Setup.RT=2.4;
CONN_x.Setup.nsubjects=1;
CONN_x.Setup.voxelresolution=4;
CONN_x.Setup.preprocessing.steps='default_ss';
CONN_x.Setup.preprocessing.sliceorder='interleaved (bottom-up)';
CONN_x.Setup.preprocessing.diffusionsteps=40;
CONN_x.Setup.done = 1;
CONN_x.Denoising.done = 1;
CONN_x.Analysis.done = 1;
CONN_x.Results.done = 1;
CONN_x.Analysis.type = 1;
CONN_x.Analysis.name='DefaultMode';
CONN_x.Analysis.sources = {'networks.DefaultMode'}
CONN_x.QA.plots = {'QA_NORM structural', 'QA_NORM functional', 'QA_NORM rois', 'QA_REG functional', 'QA_REG structural', 'QA_COREG functional', 'QA_TIME functional', 'QA_TIMEART functional', 'QA_DENOISE histogram', 'QA_DENOISE timeseries', 'QA_DENOISE FC-QC', 'QA_DENOISE scatterplot', 'QA_SPM design', 'QA_SPM contrasts' ,'QA_SPM results', 'QA_COV'}
conn_batch(CONN_x);
Where is the error?
I also tried to register the structural and the functional images (they are acquired in the same session) before run CONN but the problem has not been solved.
Thank you for your help!
Silvia
Threaded View
| Title | Author | Date |
|---|---|---|
| silvia de francesco | Mar 24, 2023 | |
| Alfonso Nieto-Castanon | Mar 29, 2023 | |
| silvia de francesco | Mar 30, 2023 | |
| silvia de francesco | Mar 31, 2023 | |