Quantcast
Channel: NITRC CONN : functional connectivity toolbox Forum: help
Viewing all 6872 articles
Browse latest View live
↧

Remove subject for ICA

June 18, 2019, 7:36 am
$
0
0
Dear Conn users,
I ran a group analysis in resting state on two populations (patients and controls). In order to see differences in network distribution, I ran ICAs on these two groups.
However, I can't find the classic DMN network. If I understand correctly, the ICAs are calculated on the overall population. So I would like to be able to run ICAs only on controls. And I have the impression that to do this I have to start a new project by running the controls, don't I? So I come to my question: how to remove subjects (which are not in a logical order... of course, otherwise it would be too simple)?

Please excuse me if my question is trivial, but I can't find the answer on the forum. But thank you in advance if you can help me!

Johanna
↧
Search

RE: strategy to analyse surface-based data

June 18, 2019, 9:58 am
$
0
0
[color=#000000]Hello[i],[/i][/color]
[color=#000000][i] [/i][/color]I just came across this old post below when I was searching through the forum.  I would like to do something very similar.  Does anyone have any hints as to how to take a significant cluster from a FreeSurfer surface-based analysis and use the cluster as an ROI in the Conn Toolbox for functional connectivity analyses?

Thank you,
Chelsea
[i]
Originally posted by Vivian Roger Steiger:[/i][quote]dear experts

do you have any new ideas to this question below?

best wishes

[i]Originally posted by Vivian Roger Steiger:[/i][quote]Dear experts,

I'm a new user of Conn and need some advice from pros on how to properly setup a surface-based analysis.

So far I've conducted a surface-based analysis with FreeSurfer, where I got a lot of mc-corrected clusters in a simple group comparison approach.
(I've used 'make_average_subject' to generate a specific analysis space for stats instead of fsaverage since I have a special population)

I'd like to use these significant mc-corrected clusters to start a seed-based analysis using CONN and the corresponding resting state data in order bring structural and functional findings together.

What would be the optimal strategy to start such an analysis?

Thank you very much for your help

Best,

vivian[/quote][/quote]
↧
↧

RE: How to transform a subset of MNI-space first-level maps to native space for each subject?

June 19, 2019, 6:33 am
$
0
0
[color=#000000]Hi,[/color]

[color=#000000]I ran into the same problem, and no matter what I try, it doesn't seem to work in SPM12. Here is what I did:[/color]

[color=#000000]1. T1 image segmented using "Segmentation" and the TPM.nii. Output is in native space, besides the c1-c5 maps I also generated the deformation fields (y_ and the inverse iy_).[/color]

[color=#000000]2. When I use "Normalise (Write)" with a label map in MNI space (e.g. labels_Neuromorphometrics.nii) and the inverse deformation field, I end up with a volume which is not really registered with the original T1 image.[/color]

Can anybody point me to some recipe how to transform MNI to native space with the deformation fields from the segmentation process?

Thanks,
Stefan
[color=#000000] 
[/color][i]Originally posted by Pravesh Parekh:[/i][quote][color=#000000]Hi,[/color]

[color=#000000]A usual rule of thumb for converting data from MNI space to native space is to use the inverse deformation field. When performing the segmentation steps, SPM will write out the forward and the reverse deformation fields (by default they are turned off; you need to select whether you want to save them or not). When running the preprocessing steps in Conn, Conn saves both these fields for you (the forward deformation is the y_*.nii and the inverse deformation is the iy_*.nii file).[/color]

You can run the normalization step (Normalize: Write) in SPM and specify the deformation field as the iy_*.nii file and select the images in MNI space (that you want to convert to native space) as the images that you want to write. Be sure to specify the correct values for voxel size and the bounding box (as per your native space file). If you set the bounding box to NaN(2,3), SPM will figure it out from the specified deformation field.

Hope that helps

Best
Pravesh
[/quote]
↧

Error when seting up 1st-level analysis

June 21, 2019, 6:04 am
$
0
0
Hello,
I have a dataset of37 subjects in a single session. The preprocessing with the default pipeline and denoising step went fine but when I try to do a seed-to-voxel or ROI-to-ROI I get an error and there are no timeseries displaying (where there is usually a graph). 

ERROR DESCRIPTION:

Index exceeds array bounds.
Error in conn_menu (line 839)
for n1=1:size(title,1),set(position.h4(1+n1),'xdata',n1+[0 0],'ydata',[0 title(n1)],'zdata',title(n1)+[1 1],'linestyle',':','marker','none','color',[.5 .5 .5],'tag','none');end
Error in conn (line 6622)
conn_menu('updateplotstack',CONN_h.menus.m_analyses_00{3},xtemp);

Has anyone ever encountered such an error? How can I fix it?

Thank you,
Olivier Roy
↧

CAT12 Surface-based ROIs as seeds in CONN analyses?

June 24, 2019, 12:13 pm
$
0
0
Has anyone here used CAT12 surface-based atlases (e.g., HCP MMP1) in CONN for functional connectivity analyses?  Is it possible to port the "lh.aparc_HCP_MMP1.freesurfer.annot" and right hemisphere atlases in CAT12 into CONN?

Thanks for any recommendations or suggestions.

Warm regards to all,
Jeff
↧
↧

How is Spatial match to template correlation value calculated?

June 25, 2019, 6:42 am
$
0
0
Hi!

To evaluate reproducibility, I compared two set of ICAs. I found information about how the Dice similarity coefficient (DSC) is calculated online, so I have an idea how that works, but there is another option in conn, calculation of the 'correlation' which yields an r value. But I found no information about how this is calculated in the manuals.
Is it e.g. calculated form thresholded whole-brain voxel z-scores of the functional maps? Can I read about this in some kind of documentation?

Thanks in advance
↧

CONN 18B error: duplicated ROI name

June 27, 2019, 10:13 am
$
0
0
Hi Alfonso and forum,

I am running 50 subjects in parallel using a batch script. I keep on receiving the error below. I saw there was a patch provided for an older version of conn (at https://www.nitrc.org/forum/message.php?msg_id=19218), but I'm not sure how useful it would be.

ERROR DESCRIPTION:

Error using conn_process (line 880)
duplicated ROi name atlas.FP r (Frontal Pole Right)

Error in conn_process (line 16)
  case 'setup', conn_disp([CONN: RUNNING SETUP STEP']); conn_process{[0:4,4.5,5]);

Error in conn_jobmanager (line 783)
  conn_process(job(n).fcn,job(n).args{2:end});

Do you know of any solutions?

Thank you in advance,

Georgia
↧

parallel computing toolbox for cluster processing

June 29, 2019, 5:59 am
$
0
0
Hi - 

Has anyone had success configuring the Matlab Parallel Computing Toolbox with a local cluster to parfor loop conn processes? Using the background/bash option (32 cores, 16-94 threads, 94 subjects - two visits) doesn't utilize the machine's full processing power for denoising (it does for QA plots). Total CPU usage with 32 threads barely > 10% at any given time. Nothing else is running simultaneously. Perhaps I have the conn HPC options configured incorrectly (default settings), so any help to speed things up will be greatly appreciated. 

-Patrick
↧
Search

ICA Metrics correlation analysis

June 29, 2019, 10:11 am
$
0
0
Hi,

I am really sorry if the answer to my question is obvious but I can’t find a clear answer in the previous studies.
I have 2 groups with resting state MRI (the first one of control, the second one of patients) that I compared using ICA with Conn. I want to correlate ICA metrics from my patients and controls to some clinical measurements (value of neuropsychological tests, survival...).
However, I don’t know what is the best value to extract from the second level analysis ICA in order to do this correlation (a « mean connectivity » for a certain network for each patient and control for example).
How could I extract such a metrics from Conn?

Thank you so much for your help!

Best,
Mickael
↧
↧

different TR for anatomical and functional images

June 30, 2019, 11:45 pm
$
0
0
Hello,

I'm working on [url=http://fcon_1000.projects.nitrc.org/indi/abide/abide_II.html]ABIDE II[/url] (UCLA Longitudinal) images. 
As mentioned in scan parameters (the attached file), there seems to be different TRs for anatomical and rest images.
How can I define this in conn toolbox? Because as I have figured out (if not mistaken) in the basic setup part, TR is being set according to the number of subjects I enter, and there is one TR for each subject.

Thanks in advance.
↧

How to generate functional connectivity matrices after first-level analysis?

July 1, 2019, 3:48 am
$
0
0
Hi all,

I have a single subject data(T1w and rs-fMRI BOLD). I am wanting to generate functional connecitivty matrix after preprocessing the raw fMRI. I already used CONN toolbox to do the entire preprocessing. I have now completed the first-level analysis as well. 

Since I am wanting to calculate FC matrix, I have computed ROI-to-ROI connectivty during first-level analysis. I belieive this is the correct way forward. I am having the output file named "resultsROI_Subject###_Condition###.mat". Can someone please show me how to compute FC matrix values using this file.
↧

RE: FIR task regression

July 1, 2019, 12:12 pm
$
0
0
Dear Alfonso,
Thank you for the guide to creating first level covariates for the FIR time delay. I'm exploring this method and everything went fine through denoising, but what settings would you recommend for the first level analysis? Previously, for event-based designs, I've used a weighted GLM with bivariate correlations and hrf weighting. However, since I've now run denoising using the FIR method, in order to take advantage of Cole et al.'s findings, would it make more sense to use no weighting in the analysis?

Thanks for your advice.

Best,
Jeff
↧

REX for CONN data

July 1, 2019, 12:30 pm
$
0
0
Hi All -

I am using REX to extract values from my CONN data. One my ROIs that I used in the analysis was taken from the "networks" that come with CONN. When I tried to load "networks" into the REX GUI, MATLAB gives me this error message:

Expected one output from a curly brace or dot indexing expression, but there were 8 results.

Error in rex>rex_image (line 1047)
[gridx,gridy,gridz]=ndgrid(1:a.dim(1),1:a.dim(2),1:a.dim(3));

Error in rex>rex_do (line 596)
[XYZMM{r},XYZWW{r},XYZNN{r},XYZnames{r},ROIA{r},ROIB{r}]=rex_image(roi_path,params.level,'image',params.select_clusters,params.selected_clusters,params.mindist,params.maxpeak,params.dims(min(r,length(params.dims))),params.roi_threshold);

Error in rex>rex_gui (line 475)
[data.params.ROIdata,data.params.ROInames,data.params.ROIinfo.basis,data.params.ROIinfo.voxels,data.params.ROIinfo.files,data.params.ROIinfo.select,data.params.ROIinfo.trans]=rex_do(data,~data.params.gui||(isfield(data.params,'steps')&&any(strcmp(data.params.steps,'nodisplay'))));

Error while evaluating UIControl Callback.

Can you please help?

Thanks,

Alan Francis
MGH/ Martinos Center
↧
↧

RE: CONN denoising & Eklund clusterwise inflation

July 1, 2019, 5:25 pm
$
0
0
Hello there, just a quick update given new development, it seems that aCompCor might not only reduce the issue but fix it altogether, this is what is suggested in the followup paper of Eklund et al, Cluster Failure Revisited, where they tested ICA but not PCA regression of noise but they expect a similar enhancement to the false positive rate inflation :-)

* Eklund, A., Knutsson, H., & Nichols, T. E. (2019). Cluster failure revisited: Impact of first level design and physiological noise on cluster false positive rates. Human brain mapping, 40(7), 2017-2032.

Also anothe paper shown that modelling accurately the autocorrelation function greatly increases fMRI reliability, which suggests the fp inflation is mainly due to the autocorrelation function being either incorrectly modelled in some software packages and/or not enough regressed.

* Olszowy, W., Aston, J., Rua, C., & Williams, G. B. (2019). Accurate autocorrelation modeling substantially improves fMRI reliability. Nature communications, 10(1), 1220.

Best regards,
Stephen
↧

Covariate per run

July 2, 2019, 3:16 am
$
0
0
Dear CONN-experts,

short question: is it somehow possible to include a covariate per run for every subject? For example something like a vigilanz estimation, that was reported from the subjects after every run.

Thanks in advance!

Iris
↧
Search

sessions comparisons

July 2, 2019, 5:04 am
$
0
0
My project is contain one healthy group and I have 2 fMRI scans (each scan has 2 sessions, one scan pre-training and the second scan post- training) . I want to know the different in connectivity between pre and post training for each subject and across the group. 
when I upload the data I upload it as 4 sessions per subject. Is that right? 
1- how to compare between sessions (2 sessions pre- and 2 post- training)?  and is it possible to compare across same subject and across the group ?
3- Is paired t test the correct choose for 2nd level analysis? 
could you please describe the way in details for me .

Many thanks
↧

Conn/Matlab keeps crashing

July 2, 2019, 9:31 am
$
0
0
Hello all,

I have a large dataset (n=229) and while conn is loading the .dmat files, it crashes around participant 200. I've tried multiple things, restarting the project from scratch, deleting the higher .dmat files, etc. The project is stuck at the first level analysis and if I delete all the .dmat files, redo the first level analysis again, it crashes again, and won't load. Help! 
Thanks!
↧
↧

CONN functional Segmentation for non-human data?

July 3, 2019, 1:05 am
$
0
0
Hi!
I'm using the CONN toolbox to analyse some rodent fMRI data, and would like to segment the individual participants into gray matter. white matter and CSF. I am wondering if the option during the preprocessing pipeline of 'Functional direct segmentation & Normalization' will work for rodent data, or if it takes the human MNI as a default input to base the segmentation on?

I have imported rodent white matter, grey matter and CSF masks in the ROI section, along with the rodent atlas I am using.

Thanks for your help!

Annelene
↧

Exporting thresholded connectivity maps

July 3, 2019, 4:26 am
$
0
0
Dear CONN users,

I have been using CONN for a month and have a question.

I am doing seed-based connectivity analyses on resting state session acquired between stimulation sessions, to assess connectivity between different regions activated or deactivated by the experimental stimulation.

I would like to be able to export the connectivity maps so I can compare them to the (de)activation maps in a viewer like fsleyes.

However, I have some trouble using the "save as" when viewing the second-level results in SPM through the results explorer.

I want to view my results at voxel-level threshold p<.05 (FDR) and cluster-level thresholding at peak voxel p<.05 (FWE) or at least cluster size p<.05 (FWE), to get fairly conservative results.

It is also important for me to be able to distinguish positive and negative effects.

However, when I view the results in SPM, the available contrast is "Connectivity result", and I can choose between FWE or uncorrected. I assume this is voxel-level thresholding.

The FWE result through SPM is far more conservative than what I can see in the Results Explorer with the settings listed above, and removes most of the interesting results.

I would appreciate any help.

Sincerely,

Andreas
Psychology student
↧

RE: sessions comparisons

July 4, 2019, 1:49 pm
$
0
0
Dear Anod,

In CONN you have the possibility to do 3rd-level analysis in the 2nd-level tab, by doing t-tests both at the sessions level and at the group level. So in your case I would input only 2 sessions per subject and setup the sessions accordingly in the "Conditions" sub-tab of the Setup tab, and create two groups (as 2nd-level covariates in the Setup tab), so that you will be able to do a 3rd-level analysis in the end as I wrote above.

I hope this helps,

Best regards,
Stephen
↧
Viewing all 6872 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>