Seed Mask

December 17, 2019, 6:59 am

≪ Previous: Slice Timing Correction in SPM for Multiband Acquisition

Hello,

I am doing a seed to voxel analysis. I would like later to export a mask that include only the Seed itself. Is there is a way of doing it (I would like to do later Intra Class Correlation and exlude later the values of the seed)?

Thank you

↧

Bug report: conn_batch

December 17, 2019, 11:44 pm

≫ Next: RE: Set up error (Any idea on how to fix this?)

≪ Previous: Seed Mask

Line 939:
if isfield(batch.Setup.subjects,'descrip'), CONN_x.Setup.l2covariates.descrip{nl2covariates}=batch.Setup.subjects.descrip{neffect};

The last variable, neffect, should be ngroup on this line (it looks like copy-paste error from later line 955)

↧

RE: Set up error (Any idea on how to fix this?)

December 18, 2019, 4:42 am

≫ Next: Defining descriptions for L2 covariates (subject effects and groups) - error in docs

≪ Previous: Bug report: conn_batch

Hi All,

Ive got a similar problem as this post, however, it seems that my spm is looking for a file i dont have.
I am trying to run a preprocessing script from somebody else in my pc and it seems it is looking for a file-path that doesnt exist on my pc. (was looking for a script that does all normal fmri preprocessing in 1 go) Does anybody have experience with this problem? Thanks in advance!

ERROR:

18-Dec-2019 12:24:15 - Failed 'Realign & Unwarp'
Error using spm_vol>spm_vol_hdr (line 80)
File "E:\MRI\BECi_Study\Data\Subject_02\Sess1\fUniS-0002-00002-000002-01.img" does not exist.
In file "C:\matlab\spm12\spm_vol.m" (v5958), function "spm_vol_hdr" at line 80.
In file "C:\matlab\spm12\spm_vol.m" (v5958), function "spm_vol" at line 61.
In file "C:\matlab\spm12\spm_realign.m" (v7141), function "spm_realign" at line 121.
In file "C:\matlab\spm12\config\spm_run_realignunwarp.m" (v6554), function "spm_run_realignunwarp" at line 78.

The following modules did not run:
Failed: Realign & Unwarp
Error using MATLABbatch system
Job execution failed. The full log of this run can be found in MATLAB command window, starting with the lines (look for the line
showing the exact #job as displayed in this error message)

↧

Defining descriptions for L2 covariates (subject effects and groups) - error in docs

December 19, 2019, 7:08 am

≫ Next: Using denoising outputs in another project

≪ Previous: RE: Set up error (Any idea on how to fix this?)

In the excellent 'conn_batch.help.txt' that provides documentation for building CONN batch specification structures, there are two entries for the 'subjects' field (lines 171 - 188), denoting effects and groups. Under both effects and groups, the same subfield 'descrip' is defined.[quote]
descrip : (optional) subjects.descrip{neffect} char array of group description (long name; for display purposes only)[/quote]
It even has the same value 'neffect' for both effects and groups (i.e. there is a typo). The typo is trivial, but what I don't know is how the field 'descrip' should be specified to be read separately for effects and groups?

The attached image shows the result of defining descrip values for 4 covariates: the first two values are also displayed for groups Patient and Control - an obvious error.

↧

Using denoising outputs in another project

December 19, 2019, 1:33 pm

≫ Next: error in "Importing functional data" step

≪ Previous: Defining descriptions for L2 covariates (subject effects and groups) - error in docs

Hello All,

I would like to import the outputs from the denoising step into another project. Is there a recommended way to accomplish this?

From,

Humza Ahmed

↧

error in "Importing functional data" step

December 21, 2019, 12:25 am

≫ Next: 2nd level voxel-to-voxel group_MVPA interpretation

≪ Previous: Using denoising outputs in another project

Hello,

I am doing the First level analysis using my task related functional MRI data set.
I did preprocessing using my own pipeline.
I have implemented conn analysis successfully using different data set preprocessed with the same pipeline.

This time, when I started First level analysis , in the step 4/7 Importing functional data, I got the error "conn_write_slice (line 18), error writing to file"
What dose this error mean ? and How should I fix this error ?
I attached the screen shot of the error message.

Thank you very much.

↧

2nd level voxel-to-voxel group_MVPA interpretation

December 22, 2019, 7:27 pm

≫ Next: Is preprocessing a deterministic process?

≪ Previous: error in "Importing functional data" step

Hello,

I am running 2nd level analysis on group_MVPA to clustering brain activation at the whole brain level.When we perform MVPA and analyze task-state fmri data, the clustering difference between condition A and condition B, I have two questions:

1. I don't know how CONN is calculated and what the added covariates represent.
2. How to visualize group results (eg. searchlight)?

I really appreciate your help.
thank you

↧

Is preprocessing a deterministic process?

December 23, 2019, 8:26 am

≫ Next: Analysing specific frequency bands

≪ Previous: 2nd level voxel-to-voxel group_MVPA interpretation

Hi Alfonso,

I have a question about preprocessing.

A year ago, I used my raw nifti data (e.g. ABCD.nii) and I ran the default preprocessing pipeline in CONN. This generated the uABCD (realigned & unwarped), the auABCD (slice time corrected), the wauABCD (segmented & normalised) and the swauABCD (smoothed).

I recently reran the default preprocessing pipeline using again the raw nifti data. At my second level analysis, I noticed some differences from my original analysis so I tried to track down where this difference comes from. To do this, I checked the intensity of a voxel across the images between the two preprocessing projects.

The original raw data (ABCD.nii) were the same between the two projects.
The realigned/unwarped data (uABCD.nii) were the same between the two projects.
The slice-time corrected (auABCD.nii) were the same between the two projects.
The segmented (wauABCD.nii) [u]were NOT the same[/u] between the two projects: the same voxel had intensity 5116.88 in one project and 5116.62 in the other project.
Obviously, differences were observed in the smoothed images, in the beta images, in the SPM.mat etc and I assume that this is why the second level analysis differs as well.

So I would like to ask
1) Is segmentation & normalization a deterministic process? Is it expected to get different results from execution to execution?
2) Is there any way to ensure the production of the exact same images after segmentation?
3) Was any change between CONN17 and CONN18 with respect to segmentation that could explain the difference?

I would really appreciate your help!

Thank you very much,
Konstantina

↧

Analysing specific frequency bands

December 23, 2019, 12:01 pm

≫ Next: Adding QC covariates to second level

≪ Previous: Is preprocessing a deterministic process?

Hello,

I am thoroughly enjoying using Conn for my analysis. Thank you for the software!

I am interested in decomposing my fMRI time series into different frequency bands and studying connectivity within each frequency band. Is there any way to do this? Also, is it possible to export (from Conn) or have as a separate output file, the time series within each frequency band?

Many thanks for your help!

↧

Adding QC covariates to second level

December 26, 2019, 9:01 am

≫ Next: 2nd lvl analysis

≪ Previous: Analysing specific frequency bands

Hi All,

I have used conn to preprocess, denoise, and run first level analyses. The preprocessing / denoising generated the usual second level QC covariates. However, I only see the AllSubject's covariate in the second level tab of the gui. I have tried to run the setup portion again, but still do not see the variables appear in the second level. I am using conn 18b.

Thanks,

Humza

↧

2nd lvl analysis

January 2, 2020, 9:01 am

≫ Next: Two resting state scans -- averaged?

≪ Previous: Adding QC covariates to second level

Hi,

I´m comparing two groups (patients and controls) and perform ROI-to-ROI analysis. When choosing "any effect among controls or patients" I have significant values. What I don´t understand is why I have no significant results when I choose "patients > controls" or "Controls > patients".

For example, I´d like to know the connectivity between PCC and thalamus in patients, and in controls, and if the difference is significant between the two groups. How do I perform this analysis? How do I know which group have the highest connectivity values? And what are the specific statistical test that is used of each of these options?

I´d also like to know the confidence interval in numbers for better understanding of the difference between the groups patients and controls.

Thank you in advance!
Hanna

↧

Two resting state scans -- averaged?

January 2, 2020, 11:27 am

≫ Next: Multiple tests for same groups

≪ Previous: 2nd lvl analysis

Hi,

I have two five-minute resting state scans for each participant. When I put them both into conn, after the preprocessing is complete, I only have 1 rest condition. To confirm, are these scans being averaged? Thanks.

Jill

↧

Multiple tests for same groups

January 5, 2020, 7:02 am

≫ Next: Error : covariates 2nd level - Index exceeds number of array elements

≪ Previous: Two resting state scans -- averaged?

Hi Afonso and all,

I have a group of subjects (20). I would like to make this group of subjects 4 groups (such that those same 20 subjects are going to be in 4 groups but they differ in something in their analysis). For example I would like to test for changes in their analysis data.

So:

Grouop 1 (20 subjects): original
Group 2 (same 20 subjects): analysis 1
Group 3 (same 20 subjects): analysis 3
Group 4 (same 20 subjects): analysis 4

My question is what is the best way of entering this data and statistically testing them?

Also can I do different set of preprocessing analysis for each group in the same CONN model file?

Thank you

Aser

↧

Error : covariates 2nd level - Index exceeds number of array elements

January 7, 2020, 7:58 am

≫ Next: Best approach for pre/post w/ possible perfusion changes?

≪ Previous: Multiple tests for same groups

Dear all,

When i try to provide different covariates for 2nd level analysis , i am incurring an error which suggests index exceeds number of array elements

I have counted the array index and it exactly matches the number of subjects that i have .

I have few functional files where number of functional slices are irregular, for 20 subjects i have 250 slices and for the rest of the subjects i have 300 slices. does this cause the error described below.

Index exceeds the number of array elements (100).
Error in conn_menu (line 839)
Error in conn (line 4238)
CONN18.b
SPM compiled
Matlab v.2018b
project: CONN18.b
storage: 310.7Gb available

ans =
logical
1
ERROR DESCRIPTION:

Index exceeds the number of array elements (100).
Error in conn_menu (line 839)
Error in conn (line 4238)
CONN18.b
SPM compiled
Matlab v.2018b
project: CONN18.b
storage: 310.7Gb available

ans =
logical
1
ERROR DESCRIPTION:

↧

Best approach for pre/post w/ possible perfusion changes?

January 8, 2020, 7:41 am

≫ Next: SPM12 prepro batch implementation

≪ Previous: Error : covariates 2nd level - Index exceeds number of array elements

I would like solicit ideas on what functional connectivity processing methods/corrections might be best when one has pre/post data where the likelihood of perfusion changes over the scan and re-scan interval may be significant (e.g., aerobic exercise intervention, increase cardiac output, etc.).

Global signal regression comes to mind, but doing so runs the risk of producing illusory anti-correlations. But, what about using the global signal QA variable as a covariate?

I've noticed that a few of the CONN voxel-wise metric analyses (e.g., ICC, ILC, MVPA) have normalization options and have wondered if the normalization process would reasonably correct for any perfusion differences contributing to pre/post FC change analyses?

The ROI-to-ROI analyses in CONN are also expressed as normalized Z, but I'm not certain if this is equivalent to the normalization options for some of the voxel-wise metrics?

Any other thoughts or recommendations on how to reasonably address regional or global perfusion differences in FC analyses, as implemented in CONN?

Warm regards,
Jeff

↧

SPM12 prepro batch implementation

January 9, 2020, 5:12 am

≫ Next: What is the contrast definition at 2nd level for comparing the Condition A of control group with Condition B of Patient group?

≪ Previous: Best approach for pre/post w/ possible perfusion changes?

Dear Alfonso,

We are trying to apply identical preprocessing pipeline of SPM12 (located at "...SPM12\batches\prepro_fmri.m") via CONN.
However, at the coregistration stage, CONN doesn't allow us to select skull stripped structural image instead of original.
Do you have any advise for us?
Thank you so much for your help.

Kind Regards
Ali Bayram

↧

What is the contrast definition at 2nd level for comparing the Condition A of control group with Condition B of Patient group?

January 9, 2020, 7:15 am

≫ Next: CONN Preprocessing error

≪ Previous: SPM12 prepro batch implementation

Dear All,

We have a study including 1 session for control group and 2 sessions(pre and post treatment) for Patient group.
We want to compare functional connectivities (seed to voxel) among those three dataset (control-prePatient, control-postPatient, prePatient-postPatient).

CONN allow us to define 1 or 2 sessions for a subject (with session invariant structural image). We have no problem for defining groups and sessions.

We defined 2 conditions: preRest & postRest for comparison in the second level anaysis.
One session of Control group defined for both conditions, while two sessions of Patient group defined separately for conditions.

But we have problem while adding effect of rest at the denosing step. There is double "Effect of rest" for control group (pre and postRest) due to the our condition definition style.

Do you have any advise for this type of comparisons?

Is there a contrast definition at the second level for comparing control(preRest) vs. patient(postRest)?

Best Regards.
Ali

↧

CONN Preprocessing error

January 10, 2020, 11:24 am

≫ Next: how to employ func.gii to perform surface-based FC analysis?

≪ Previous: What is the contrast definition at 2nd level for comparing the Condition A of control group with Condition B of Patient group?

Hello CONN community,

I am currently running my preprocessing step on CONN and kept meeting with this error.

[b]Computing mask... : Failed 'Realign & Unwarp'[/b]
[b]Matrix dimensions must agree.[/b]
[b]In file "/space/-/-/spm/spm12/spm_uw_apply.m" (v6301), function "spm_uw_apply" at line 229.[/b]
[b]In file "/space/-/-/packages/spm/spm12/config/spm_run_realignunwarp.m" (v6554), function "spm_run_realignunwarp"[/b]

I have about 100+ subjects and a very few of them have this issue. The preprocessing step would continue running for other participants, but it crashes before going on to the next preporcessing step. Some participants have less sessions than others, would this be the issue?

Can I please get some help on this?

Thank you.

↧

how to employ func.gii to perform surface-based FC analysis?

January 10, 2020, 8:04 pm

≫ Next: Error when working with Conditions

≪ Previous: CONN Preprocessing error

Hi everyone,

To conduct surface-based FC analysis, the manual shows that we should enter the T1 or brainmask volumes in the subject-specific mri folder generated by Freesurfer, but it still uses the voxel-based functional images.

Here, I conducted surface-based preprocessing using fMRIPrep, and obtained L/R func.gii file for each task session. How can I use these files to perform surface-based FC analysis in conn?

Thanks in advance!

↧

Error when working with Conditions

January 12, 2020, 4:40 pm

≫ Next: RE: Implementation of graph theory in task fMRI

≪ Previous: how to employ func.gii to perform surface-based FC analysis?

Any changes I make in the Conditions results in an error. This happens when I try to create a condition manually, or when I try to import my .txt file.
Thanks for any help.

[code]ERROR DESCRIPTION:[/code][code]Error using reshape
Size vector must have at least two elements.[/code][code]Error in conn_convertcondition2covariate (line 63)
x=mean(reshape(x(offs+(1:10*CONN_x.Setup.nscans{nsub}{nses})),[10,CONN_x.Setup.nscans{nsub}{nses}]),1)';%x=x(1+10*(0:CONN_x.Setup.nscans{nsub}{nses}-1));[/code][code]Error in conn (line 3594)
out=conn_convertcondition2covariate('-DONOTAPPLYSUBJECTS',tnsubs,1:numel(CONN_x.Setup.conditions.names)-1);
CONN18.b
SPM12 + Anatomy DEM FieldMap MEEGtools bspmview marsbar wfupickatlas
Matlab v.2019b
project: CONN18.b
storage: 354.9Gb available[/code][code]spm @ /Volumes/Data/Drive/MATLAB/spm12
conn @ /Volumes/Data/Drive/MATLAB/conn[/code]

↧