Quantcast
Channel: NITRC CONN : functional connectivity toolbox Forum: help
Viewing all 6871 articles
Browse latest View live
↧

preprocessing and adding ROIs in batch

December 1, 2016, 2:57 pm
$
0
0
Hi Alfonso,

I have a batch script (attached, adapted from your HCP example script) that ran well in my initial tests, and now I'd like to add some additional ROIs. I see from a previous thread (https://www.nitrc.org/forum/forum.php?thread_id=5822&forum_id=1144) that there is an option for adding new ROIs in batch. However, when I added new ROIs to my list, set the flag batch.Setup.rois.add=1, and run the batch script, it begins rerunning some of my preprocessing steps (segmentation and ART-repair, though not smoothing). Preprocessing is fairly time/resource intensive and I plan to periodically add new ROIs and re-run this model, so I'd like to avoid this if possible. Commenting out the preprocessing options allows Setup to run to completion but causes other issues (the loaded functional images revert to the unsmoothed versions and the scrubbing first-level regressor disappears, causing the Denoising step to crash; the segmentation data still loads correctly). Is there a simple way to tell CONN to skip preprocessing but still load the preprocessing outputs when re-running (akin to setting the overwrite option to "No" for other steps)? I'm using version 17a.

Thank you!
-Ely
↧
Search

Looking at the output from explorer

December 1, 2016, 5:29 pm
$
0
0
Hi everyone,

The question is more a sanity check to be sure I understand the output since it's the first time I'm looking at the output from CONN. 

I'm currently looking at the second level analysis for a 2x2 mixed design with resting state scans where each subject has 2 sessions (each 4 weeks apart) and I done the following:
<span style="white-space: pre;"> </span>- Conditions have been designed with pre and post where the pre condition session 1 is defined as Onset = 0 and Duration = inf while session 2 is Onset = [] and Duration = [] and opposite for the post condition. 
<span style="white-space: pre;"> </span>- Under Second-level covariates I have added two called group_1 and group_2 that are defined as follows (example with 8 subjects): 
<span style="white-space: pre;"> </span>- group_1 = [1 1 1 1 0 0 0 0];
<span style="white-space: pre;"> </span>- group_2 = [0 0 0 0 -1 -1 -1 -1];

When I get to second level analysis I marked both group_1 and group_2 under subject effects and pre and post under conditions and below I have added the following:
<span style="white-space: pre;"> </span>- for Between-subject contrast = [1 -1]
<span style="white-space: pre;"> </span>- for Between-conditions contrast = [1 -1] (Pre > Post)

[b]Is this the right way to input variables? And if I get positive values (red lines) does that mean that post is bigger than pre? [/b]

Thanks a lot in advance. 

Best,

Eric
↧
↧

customize roi

December 2, 2016, 8:17 am
$
0
0
Hi Alfonso & group!

I would like to ask whether there is a way to work in conn with customized rois as seed regions?

Best,
Isadora
↧

Importing ICA components to Conn

December 4, 2016, 9:13 am
$
0
0
Hi Alfonso

I used the ICA GIFT to determine components for groups and sessions according to default setting (infomax, z-score, etc.). How (if it is easily done) can I import them to run the 2nd level statistics in Conn? Is there still anything to consider regarding compatibility?

I would greatly appreciate your suggestion
Kind regards,
Lucas
↧

RE: Error in preprocessing (realign)

December 4, 2016, 6:48 pm
$
0
0
Hi Olaia and Alfonso,

I am also wondering what you mean by faulty images? I have EXACTLY the same error (both error message that I am getting and also the error described in the Command Windows) as Olaia. I am not sure where this error is coming from and what about the functional images should be checked. Thank you!

SPM12: spm_uw_apply (v6301) 21:13:08 - 04/12/2016
========================================================================
Computing mask... : ...done
Reslicing images... : ...done
Writing mean image... : ...done
Completed : 21:16:40 - 04/12/2016
Done 'Realign & Unwarp'
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)
------------------
Running job #1
------------------
CONN v.16.h
SPM12 + DEM FieldMap MEEGtools
Matlab v.2016b
↧
↧

Would you mind helping me? CONN Beginner

December 5, 2016, 9:20 am
$
0
0
Dear Sir and Madam

I want to use CONN to show the difference in functional connectivity between patient and healthy group, then I have prepared one patient data(.nii format) and 10 healthy control data(.nii format).
I have set up those 11 people data in Basic, Structural, and Functional Data Session. then I started CONN.
After preprocessing (No QA error), I confirmed the covariates 1st level, and set up covariates 2nd level as follows:control as 0, patient as 1.
Then I set the "Option" setting as default condition(checked all boxes; Roi-to-Roi, etc.). Then I did denoising.
After that I finally got the analysed result. But I can't understand what does this result mean?
I want to know how to depict  the difference in functional connectivity between patient and healthy group.
I am struggling this problem. so please teach me and mail me below.

Best Regard.

Tsuyoshi Hashimoto PhD.
Advanced Imaging Center Tsukuba JAPAN
E-mail: ytm.workspace@gmail.com
↧

Updating reference

December 5, 2016, 9:20 am
$
0
0
Hello,

I am trying to load my conn project. When the program checks if the data has been moved, I'm prompted for a .nii file. I have not moved the data though. When I select a file from the original data directory, the window pops right back up and asks me again for a .nii file. This is repeated every time I select a file. When I select "cancel", none of the functional or structural files are present in the project. 

The pre-processing step already ran so the processed files should be in there, which would make it more difficult to upload them again. Plus, I have 460 files (9 functional sessions + structural for 46 participants), so it would really be great if I didn't have to re-upload all of those files and it potentially happening again!

Any tips would be great, thanks!
↧

Extracting voxel time-series

December 5, 2016, 10:36 am
$
0
0
Dear CONN users, Dear Mr. Nieto-Castanon,


I would like to extract time series for each of my subjects by using anatomical ROIs given in the tool (DMN related) and from the ICA maps I retrieved. How could I do this and which files shall I use for this process? I would greatly appreciate any information and suggestion you can provide me with. Thank you for your time.


Best regards,
Josephine
↧
Search

time courses from cut-off areas

December 6, 2016, 11:53 am
$
0
0
Hello!
This is more an analysis-related question: I have a within-subjects design with three measures per subject. In some cases, the coverage of the postcentral/precentral gyrus is incomplete (aprox 20% uncovered after discharged very affected images). I extracted the time courses of the images, based on the HO atlas and I want to proceed with the graph analysis. However, I am concerned about this issue, because the correlation coefficients of the cut areas are for sure, affected in some participants or in some sessions. I am thinking that the results will be biased due to this issue.
Do you have a suggestion? Should I exclude the whole area?
I am looking forward to your suggestions.
Thanks
YIDA
↧
↧

RE: error messages throughout pipeline

December 6, 2016, 1:21 pm
$
0
0
[color=#000000]Hi Eran,[/color]

[color=#000000]I believe this error simply indicates that you have not yet named your CONN project file. Clicking on 'save' and giving your project a name should fix this issue.[/color]
[color=#000000]Best[/color]
[color=#000000]Alfonso[/color]
[i]Originally posted by Eran Dayan:[/i][quote]Hi,

I recently upgraded my Conn version (to 16B, with SPM 12 on MATLAB 2016b, all running on OS X El Capitan).  I keep getting the following error messages, while preprocesssing the data, although, for the most part the program appears to execute most of the steps in the pipeline correctly. Any idea what's going on?

Thanks!

Eran

ERROR DESCRIPTION:

Error using conn (line 677)
invalid filename, project NOT saved
Error in conn_setup_preproc (line 1362)
conn save;
Error in conn (line 776)
ok=conn_setup_preproc('',varargin{2:end});
Error in conn_menumanager (line 119)
feval(CONN_MM.MENU{n0}.callback{n1}{1},CONN_MM.MENU{n0}.callback{n1}{2:end});
CONN v.16.b
SPM12 + DEM FieldMap MEEGtools suit
Matlab v.2016b
Warning: Contents.m overloaded by version in folder /Applications/MATLAB_R2016b.app/toolbox/matlab/datafun
Warning: spm_cfg.m overloaded by version in folder[/quote]
↧

RE: connectivity in CSF

December 6, 2016, 1:32 pm
$
0
0
[color=#000000]Hi Elsa,[/color]

A few things to check could be:

a) make sure all subjects have been correctly normalized to MNI space; you may do that for example in the [i]Setup.Functional[/i] tab by clicking on 'functional tools. check functional-MNI coregistration' (see also 'check functional-anatomical coregistration' and 'display same slice for all subjects' functions there)

b) make sure that the White/CSF masks look correctly coregistered/normalized as well; for example in the [i]Setup.ROIs[/i] tab, select the White Matter ROI and click on 'roi tools. check ROI-functional coregistration'

and c) make sure that both the WM and CSF effects are included in the [i]Denoising [/i]tab under the 'confounds' list (and that the number of components/dimensions used for each of these ROIs is sufficiently high, at least 3 or higher)
 
If all of those look fine, then, perhaps other things to consider would be whether, depending on the threshold that you are using to display your first-level connectivity maps, perhaps these are just spurious/random correlations (are these consistent across subjects?) and/or whether perhaps they are located not purely in CSF areas but in nearby subcortical areas?

Hope this helps
Alfonso 

[i]Originally posted by Elsa Ermer:[/i][quote]Hi,

I'm running a conn analysis on a group of stroke patients and ending up with some subjects who have a fair amount of connectivity in CSF in the first-level analysis ROI seed-to-voxel results. I am following the default settings for the analysis and the ROIs for CSF, WM etc in the Setup look reasonable. Does anyone have any suggestions for errors I might be making?

Thanks,
Elsa[/quote]
↧

RE: Conn17a Networks

December 6, 2016, 2:43 pm
$
0
0
I too am curious about this. Since it seems the old "DMN" option was subsumed in the newest version of CONN under the new "networks" heading, I am wondering which nodes should be considered for DMN connectivity analyses.
↧

Multiple comparison correction question

December 7, 2016, 2:31 am
$
0
0
Dear Conn experts, 
I am new to Conn and I have got question regarding multiple comparison correction. I am comparing two groups. If I select all ROIs in Results explorer , some connections in connectome ring show after FDR correction significant effect. My question is, if conn compares all possible ROIs and every possible connection between ROIs is FDR corrected, then there  is another need for multiple comparison correction? In another words if some connections show  significant effect are they real or they could be spurious because of the need for another multiple comparison. Thank you.

Boris
↧
↧

CONN and ASL

December 7, 2016, 6:08 am
$
0
0
Dear Conn-Usergroup,

We want to compare BOLD resting state with ASL. We managed to run the BOLD data through CONN and the results seem resonable. For ASL we threw in the difference images (label vs control). The preporcessing showed good results but when it comes to denoising it seems that all the variance is in the difference images is categorized as noise, especially by the realignment confound. First and second level analysis results do, with a lot of good will, have a tendency towards the BOLD analysis but are not really sensible. Does anyone have experience with using CONN to analyse ASL data?

Cheers
Martin
↧

Longitudinal study

December 7, 2016, 7:54 am
$
0
0
Hi,
we conducted a longitudinal resting-state FC study and we would like to use CONN for the analysis. We already preprocessed (canonical + rs) our data.
We have one group of healthy subjects, studied in three sessions (each with 6 runs of resting state condition) in three different days. We would like to know how to run a one-factor repeated-measures Anova with our three timepoints. many thanks
Giorgia
↧
Search

RE: Error after set-up

December 7, 2016, 2:27 pm
$
0
0
Got this same error in CONN 16b.  Denoising step freezes at step 6/6: updating analysis variables.  Get the following error in the command window:

CONN: RUNNING DENOISING STEP
Not ready to process step conn_process_9

Only voxel-to-voxel was selected for enabled analyses, no intermediate optional files are being saved.  Everything else is pretty standard!
↧

RE: conn17a ERROR: duplicate ROI labels

December 8, 2016, 8:26 am
$
0
0
Thank you for posting this. I had the same problem and the new rex.m fixed it.
↧
↧

Processing

December 8, 2016, 8:37 am
$
0
0
Hi all,
I wanted to ask what would be the first job on Matlab10 while running the processing step on CONN15h after setting the 2nd level covariates as it showed segmentation. I wanted to make sure this was right.

Thanks
↧

Preprocessing in 17a

December 8, 2016, 8:56 am
$
0
0
I'm using the GUI to preprocess some NIFTIs but am running into an error on the smoothing step. Output is below, but after searching through the forums here I've come to understand the error is a result of the meanu***.nii files having multiple volumes.

The advice given in the threads I've found suggest checking a 'first functional volume as reference' box somewhere in the preprocessing GUI, but despite my best efforts I can't seem to find it. Can someone kindly guide me to the right place in 17a?

Thanks!

[code]
Running 'Normalise: Estimate & Write'
Smoothing by 0 & 8mm..
Failed 'Normalise: Estimate & Write'
Expected one output from a curly brace or dot indexing expression, but there were 5 results.
In file "/home/joe/Documents/MATLAB/spm8/spm_smoothto8bit.m" (v4310), function "smoothto8bit" at line 33.
In file "/home/joe/Documents/MATLAB/spm8/spm_smoothto8bit.m" (v4310), function "spm_smoothto8bit" at line 16.
In file "/home/joe/Documents/MATLAB/spm8/spm_normalise.m" (v4621), function "spm_normalise" at line 153.
In file "/home/joe/Documents/MATLAB/spm8/config/spm_run_normalise_estwrite.m" (v4152), function "spm_run_normalise_estwrite" at line 22.[/code][code]The following modules did not run:
Failed: Normalise: Estimate & Write[/code]
↧

Error in Conn

December 9, 2016, 12:13 pm
$
0
0
Hi 
I have below problem in conn17
 What should I do?

ERROR DESCRIPTION:

Error using conn_process (line 763)
duplicated ROI name atlas.Ver10 (Vermis 10)
Error in conn_process (line 15)
case 'setup', disp(['CONN: RUNNING SETUP STEP']); conn_process([0:4,4.5,5]);
Error in conn (line 3195)
else conn_process('setup');
Error in conn_menumanager (line 119)
feval(CONN_MM.MENU{n0}.callback{n1}{1},CONN_MM.MENU{n0}.callback{n1}{2:end});
CONN v.17.a
SPM12 + DEM FieldMap MEEGtools
Matlab v.2015a
storage: 370.9Gb available
Warning: Contents.m overloaded by version in folder D:\Darsi\AUT\ms_project\AUT_Group_Sessions\eeglab13_6_5b
↧
Viewing all 6871 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>