NITRC CONN : functional connectivity toolbox Forum: help

RE: GLM variables and Exporting speciffic data

Abby Paulson — Wed, 28 Aug 2024 17:51:54 GMT

Error using svd SVD did not converge.

Yuan Cao — Sun, 04 Aug 2024 7:45:33 GMT

Hi expert,
I got the following error at the first level of denoising when I got to 51 objects. 
Error：
Step 5/6: preprocessing voxel-to-voxel covariance ERROR DESCRIPTION:   Error using svd SVD did not converge. Error in conn_process (line 1917)  [Q1,D]=svd(Cy); % Q1*D*Q1 = Ctt (time-by-time covariance matrix; note:unit variance timeseries) Error in conn_process (line 69)  case {'preprocessing_gui','denoising_gui'}, conn_disp(['CONN: RUNNING DENOISING STEP']); conn_process([1.5,2,6:9],varargin{:}); Error in conn (line 7024)  else conn_process('denoising_gui'); ispending=false; Error in conn_menumanager (line 124)  feval(CONN_MM.MENU{n0}.callback{n1}{1},CONN_MM.MENU{n0}.callback{n1}{2:end});  CONN22.a SPM12 + DAiSS DEM FieldMap MEEGtools Matlab v.2018a project: CONN22.a storage: 2148.7Gb available   spm @ /media/**/**/toolbox/spm12 conn @ /media/**/**/toolbox/conn22a/conn
 
Thanks,
Yuan

Issue with Loading MAT-file in CONN

Rasha Alharthi — Sat, 03 Aug 2024 22:06:21 GMT

I am experiencing a problem with loading a MAT-file in CONN during the setup process. The specific error occurs at the step where ROI data is being imported.
Error Details:
<ul>
<li>CONN Version: CONN22.a</li>
<li>SPM Version: SPM12</li>
<li>MATLAB Version: MATLAB v.2020a</li>
<li>Error Description:</li>
</ul>
Error using load Unable to read MAT-file /Volumes/My Passport/conn_project01/data/ROI_Subject103_Session001.mat. File might be corrupt.
Error in conn_process (line 853)  old=load(filename);%,'data','names','source','xyz','sampledata','samplexyz');
Error in conn_process (line 55)  case 'setup', conn_disp(['CONN: RUNNING SETUP STEP']); conn_process([0:4,4.5,5]);
Error in conn (line 6392)  conn_process(processname); ispending=false;
Error in conn_menumanager (line 124)  feval(CONN_MM.MENU{n0}.callback{n1}{1},CONN_MM.MENU{n0}.callback{n1}{2:end});
Could you please advise on how to proceed with this issue? Is there a way to repair or recover the file, or should I attempt to regenerate the file? Any guidance or suggestions would be greatly appreciated.
Thank you, Rasha

GLM variables and Exporting speciffic data

Irina Luz — Fri, 02 Aug 2024 15:52:58 GMT

Hi everyone!
I'm very new to using Conn or every other fMRI toolbox, but I have two simple (I believe) questions.
1st: After completing every preprocessing and denoising step, I ran an ICA on my data. After this, I wanted to correlate a speciffic ICA with a self-report instrument result to see any individual differences relating to that. When I try to run this analysis, Conn gives me a "warning" about having to select a "constant term" (suggesting me to use the "AllSubjects" variable created after the preprocessing). However, it seems odd to me to include this variable because I'm trying to check for differences between the individuals. Can I ignore this warning or should I really introduce the variable?
2nd: For other analysis I need to retrieve from my data the levels of activation of the chosen ICA. I tried to use the "Export Mask" option, but it gave me an image that I cannot open with FSL. Is there another way to do this? Or even export the mean activation values for excel or something like that?
Thank for your time :)
Irina Luz

Seed-Bases Sliding Window Analysis

Sofia Amaoui — Thu, 01 Aug 2024 16:05:28 GMT

Dear all, 
We are running a whole-brain dynamic functional connectivity (dFC) to explore potential brain nodes (attractors). Here’s a summary of our process and the issues we’re encountering: 1. We conducted a whole-brain dFC study using CONN for preprocessing and denoising with a specific band-pass filter. 2. We then resized the voxels to 6mm and applied the attractors code. This analysis revealed two significant clusters.
We now aim to perform a seed-based analysis using these specific ROIs, incorporating a sliding-window approach in CONN to capture the intrinsic dynamics similar to our initial whole-brain analysis. To achieve this, I have attempted to:
<ul>
<li>Add the denoised and resized brain images and skip the preprocessing and denoising steps.</li>
<li>Modify the setup section to include the sliding windows and ROIs.</li>
<li>Run the seed-based analysis as usual.</li>
</ul>
However, the results from the second-level analysis appear unusual, with all clusters showing the same size. I believe there may be an error in my approach, but I can't figure it out.
Could you please help with the following:
- Are preprocessing and denoising steps necessary for performing a sliding-window analysis?
- Should I prepare the data in a specific way?
- How should I interpret the second-level results from this analysis?
 Thank you for your assistance.
Best regards. Sofia Amaoui,

Different correlation matrices generated from the same data.

Chihhao Lien — Thu, 01 Aug 2024 15:12:23 GMT

Dear experts,
It's my first time doing longitudinal analysis with CONN.
According to previous replies to the setting for longitudinal studies, there are 2 approaches.
1:  Enter the different sessions as if they were different subjects (https://www.nitrc.org/forum/forum.php?thread_id=4955&forum_id=1144)
2: Enter 2 as the "number of sessions" per subject (https://www.nitrc.org/forum/message.php?msg_id=12664; https://www.nitrc.org/forum/message.php?msg_id=18473)
I thought that CONN would generate the same ROI-to-ROI matrices with both approaches since I extracted these matrices from the result of 1st level analysis. However, I got different correlation matrices (Fisher's Z values, extracted from "resultsROI_Subject00*_Condition00*.mat") for my follow-up data.
I tested 4 different settings:
(1) Import HC and patient data at baseline, each subject has 1 session
(2) Import baseline and follow-up data (patients only) as different subjects (e.g. import baseline data as Subject 1-10, and then import follow-up data as Subject 11-20)
(3) Import baseline and follow-up data (patients only) as 2 sessions for 1 subject. Each subject has 2 sessions: baseline and follow-up
(4) Import follow-up data only, each subject has 1 session. 
While comparing the Z matrices for baseline data, (1), (2), and (3) generated the same matrices. However, while focusing on follow-up data, (3) generated different correlation matrices, compared with (2) and (4). 
I plan to extract ROI-to-ROI matrices for other analyses (e.g. graph theory), but I', confused about which setting I should use for my longitudinal data. Does anyone know which one is the better approach?
Best regards,
Chih-Hao

Human Connectome Project (HCP) data in new version of CONN?

t_hicks — Wed, 31 Jul 2024 16:33:48 GMT

Is the Human Connectome Project (HCP) dataset included in the new FCP data provided with the newest version of CONN?

RE: Setting for longitudinal studies

Chihhao Lien — Tue, 30 Jul 2024 10:08:13 GMT

Dear experts.

I'd like to provide more information about my question.

I have 16 patient data at baseline and follow-up.

All data were preprocessed and denoised via fMRIprep and ICA-AROMA denoising.

I use "New (import)" -> "from fMRIprep dataset" to import preprocessed data into CONN, and then replace functional data with denoised data.

I tested the ROI-to-ROI matrices generated by 3 different settings.

(1) only import follow-up data
(2) import baseline and follow-up data as 2 sessions for each patient
(3) import baseline and follow-up data separately, resulting in 32 subjects in the setting, each subject has one session (baseline or follow-up)

I extracted ROI-to-ROI matrices from “resultsROI_Subject*_Condition001.mat" in "...\results\firstlevel\SBC_01" (For (2), also from “resultsROI_Subject*_Condition002.mat"). I noticed that (1) and (2) generated the same correlation matrices, but (3) generated different results.

I know (2) is the correct way to set up a longitudinal study. But, given that (1) and (3) result in different correlation matrices, I'm wondering why they generate different results. Would the same situation happen when I compare healthy controls and patients? (e.g. separately import healthy controls and patients as 2 conn project vs. import all data as one conn project).

An update: I did the same analysis as (1) again. However, the matrices are different with (2), but as the same as (3) this time.

I noticed that the second time automatically extracted more confounds (e.g. 'scrubbing_Dim20', 'QC_aroma_motion_Dim40', 'QC_w_comp_cor_Dim27'), but I only used CSF, white matter, and effect of rs for denoising, and these confounds/regressors were not included in SBC analysis, too. Thus, I don't really know why did I get different results from the same data and the same setting.

Region-to-voxel option in CONN?

Calvin Hew — Mon, 29 Jul 2024 23:58:49 GMT

Hi, sorry if it's a dumb question but from what I understand, the network we get from first level analysis is essentially Region to Region (ROI-to-ROI) matrix e.g.<code> load resultsROI_Subject001_Condition001.mat</code>
Are we able to get Region to voxel networks in a similar manner? I am asking because I am able to import these ROI-ROI matrices into python for ML. But I also want to try for Region to Voxel as suggested by a paper I read. 
If anyone could enlighten me, I'd really appreciate it! Thank you

Citation error in Write Methods tool when selecting RRC analysis only

Joseph Dust — Mon, 29 Jul 2024 20:21:31 GMT

Hello,
 I am doing MVPA, SBC, and RRC analyses in CONN and have been trying out the "Write Methods" tool to help with my Methods section. I noticed a citation error when I select only RRC for my 1st level analysis. The Group level analyses text reads as:
 "Group-level analyses were performed using a General Linear Model (GLM[12]). For each individual connection a separate GLM was estimated, with first-level connectivity measures at this connection as dependent variables (one independent sample per subject and one measurement per task or experimental condition, if applicable), and groups or other subject-level identifiers as independent variables. Connection-level hypotheses were evaluated using multivariate parametric statistics with random-effects across subjects and sample covariance estimation across multiple measurements. Inferences were performed at the level of individual clusters (groups of similar connections). Cluster-level inferences were based on parametric statistics within- and between- each pair of networks (Functional Network Connectivity[13]), with networks identified using a complete-linkage hierarchical clustering procedure[14] based on ROI-to-ROI anatomical proximity and functional similarity metrics[15]. Results were thresholded using a combination of a p < 0.05 connection-level threshold and a familywise corrected p-FDR < 0.05 cluster-level threshold[16]."
Reference [14] is reported as:
[14] Sørensen, T. (1948). A method of establishing groups of equal amplitude in plant sociology based on similarity of species and its application to analyses of the vegetation on Danish commons. Biologiske Skrifter / Kongelige Danske Videnskabernes Selskab 5: 1-34.
Luckily this citation error is removed when I select multiple 1st level analyses (e.g. MVPA, SBC, RRC) since the text changes:
"Group-level analyses were performed using a General Linear Model (GLM[13]). For each individual voxel a separate GLM was estimated, with first-level connectivity measures at this voxel as dependent variables (one independent sample per subject and one measurement per task or experimental condition, if applicable), and groups or other subject-level identifiers as independent variables. Voxel-level hypotheses were evaluated using multivariate parametric statistics with random-effects across subjects and sample covariance estimation across multiple measurements. Inferences were performed at the level of individual clusters (groups of contiguous voxels). Cluster-level inferences were based on parametric statistics from Gaussian Random Field theory[14,15]. Results were thresholded using a combination of a cluster-forming p < 0.001 voxel-level threshold, and a familywise corrected p-FDR < 0.05 cluster-size threshold[16]."
Wanted to point this out for bug reporting. I'm not quite sure why the Group level analyses text changes  when using multiple 1st level analyses compared to only RRC. It also leaves out the Jafri 2008 paper for RRC thresholds when selecting multiple 1st level analyses.