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

RE: ALFF/fALFF processing error

$
0
0
Hi,

running Conn 18.b
Same issue as above.
three groups, first group has 4 sessions, second two groups have 2 sessions each. When I run ALFF or fALFF, only the first group get results, the other two groups get all zeros.

Other voxelvise constant work fine

First time I ran setup, I forgot to indicate "allow missing data", but fixed that since, and reran setup with the overwrite option, but the fALFF and ALFF error remains

thanks
Clas

Apple cannot verify the developer for the SPM download

$
0
0
Hello!

I am having issues with Apple not allowing the Conn toolbox to use anything contained within the SPM download, due to not being able to verify the developer of SPM. Has anyone else run into this issue? How might this be resolved?

Thanks!

suspicious Denoising + Results

$
0
0
Dear all,

from the manual:
The histogram plot at the center displays the voxel-to-voxel connectivity (r values) before, and after confound removal. Typically
confounds introduce a positive bias in connectivity measures so the histogram of original connectivity values can appear "shifted" to the right. After confound removal the distribution of connectivity values appears approximately centered.



In my study (exp task and control task for 42 subjects 1 session), the original distribution and the distribution after denoising look both centered, not as explained in the manual. Running the analysis afterwards does not show any logical results. I guess this has something to do with it, but I am just new around here and I do not know what could I change.
Could somebody help me here?

Best regards
Elen

Adding a new roi to CONN

$
0
0
Dear experts,

I just started using CONN and have a question about adding a new grey matter roi.
I have a nifti file of that roi, however since it comes from a VBM analysis it is not in the right space. What would be the correct way to register this roi into the right space (space of the CONN analysis)? Does CONN have a common template I should register it to? If so, where can I find this template?

Thanks a lot in advance.

Isabelle

SMOOTHING in CONN

$
0
0
Hi all 

How does smoothing functional connectivity in CONN works? e.g. if I have a large ROI does CONN smooth the whole time series signal within this large ROI? Here I am talking about not smoothing the data but smoothing the time series signal 

Also in my analysis, I used around 7 mm kernel to smooth the raw data I.e. in the preprocessing steps. I was asked that the results might be affected by this kernel and maybe if I used a lower kernel I would not get the same results I got from this 7 mm kernel. So is there a way of answering this or testing it without the need of re analysing the whole data using a lower smoothing kernel? what if U smooth the source ROIS applied in conn and show the differences to proof that the difference is not huge. Will that works in CONN? 

Thank you 

Aser

Help with troubleshooting preprocessing error :(

$
0
0
Hi all,

I'd really appreciate your help with troubleshooting this error in the very first preprocessing step.

I'm noticing that the first level covariates of QC-timersies, realignment, and scrubbing are not being generated automatically and I don't know to fix this issue. Here is the error code, thank you for your help!

ERROR DESCRIPTION:

Error using assert
Function is not defined for 'cell' inputs.

Error in conn_loadtextfile (line 15)
assert(numel(tfields)==2,'unable to interpret filename %s',tfields);

Error in conn_getinfo (line 98)
x=conn_loadtextfile(deblank(filename(n1,:)));

Error in conn_file (line 11)
[nV,str,icon,filename]=conn_getinfo(filename,doconvert);

Error in conn_setup_preproc (line 3540)
CONN_x.Setup.l1covariates.files{nsubject}{icov}{nses}=conn_file(outputfiles{isubject}{nses}{2});

Error in conn (line 1143)
ok=conn_setup_preproc('',varargin{2:end});

Error in conn_menumanager (line 121)
feval(CONN_MM.MENU{n0}.callback{n1}{1},CONN_MM.MENU{n0}.callback{n1}{2:end});
CONN20.b

structural field and denoising questions

$
0
0
Hello!

I have already preprocessed my data and so I have skipped preprocessing in CONN. I have however a few questions:
1- For the Setup.structural, should I choose the anatomical images in subject-space or the anatomical images in MNI space? I do all my analysis in MNI space, but use the WM and CSF masks for CompCor, so I'm wondering what is the recommended approach.

2- For the denoising step, I'm a little confused what confounds I should include in the "Linear regression of confounding effects" list. Currently I'm using the defaults (WM, CSF, motion regressors, effects of rest and conditions). But is there any reason to include the effects of rest and all condition effects, or should I remove them from the list?

3- Finally, I have used the default 128s high-pass filter in SPM (but no low-pass) when setting up the 1st level model, so what values should I indicate in my band-pass filter field? Should I use the default values, or rather [0.0078 Inf] to be consistent with SPM (ie 128s high-pass filter and no low-pass filter)?

Thank you very much for your help in advance!
Sam

functional preprocessing results

$
0
0
hello, 
I have a question: 
Is there a file of the functional preprocessing that is in 4D ?
I want to see the preprocessing results on FSLeyes and I couldnt find the file.
Please let me know if there is a file in 4D what does it called? 
If not, how can I watch my preprocessing results? 

I used the preprocessing pipeline on conn. 

Thank you,
Sandy.

comparison between two sessions

$
0
0
Hi,

I hope you are all healthy and doing well. I am writing regarding setting up the group comparison between two sessions in conn.
I have two sessions of resting-state data from the same subjects.
The only thing is that session two is shorter than session one.
I want to compare seed-based functional connectivity between these two sessions. How should I set it up in the conditions tab? I have currently created two conditions: rest long and rest short. Session one has been assigned to condition long, and the interval set to the entire session. Then I have condition short assigned to session two and again with the interval set to an entire session. Then in the second-level analysis tab, I set up the contrast between conditions [-1 1]. I am not sure if the current setup is correct. I would be very grateful for any tips on how to correct it.

Cheers,
Justyna

RE: CONN compatibility with BIDS and fmriprep output

$
0
0
Hi
Just wondering if there were any updated suggestions on importing fMRIprep preprocessed output into CONN - particularly handling the confounds tsv files?

Group comparison in CONN

$
0
0
Hello,

I would like to ask whether it is possible to run the second-level group analysis with different cohort groups in CONN. As data of each group was collected separately, they have different TR, slice order information, numbers of volumes, etc. I am wondering whether I could run preprocessing, denoising, and the first level analysis separately and bring their data and compare them in the second-level analysis. Does it work the same if I get Fisher's r-to-z transformed correlation coefficients from each group and do a two-sample t-test?
Thank you for your help and time.

Best,
Su

Standalone 20b version

$
0
0
Hi:

There are any possibility to get Standalone 20b version ?

Sincerely,


Gonzalo Rojas Costa

comparison of functional connectivity (seed to voxel)

$
0
0
Hi:

  I want to compare the functional connectivity (seed to voxel) of a patient to a database of normal volunteers functional connectivity... How can I do it ?... How many healthy volunteers do I need to make the comparison?...

  Sincerely,


Gonzalo Rojas Costa

All slices shown in "Preview first-level analysis results" window

$
0
0
Hi:

  How can I view all the slices of connectivity image shown in "Preview first-level analysis results" window ?

  Sincerely,


Gonzalo Rojas Costa

RE: Adding a new roi to CONN

$
0
0
Hi:

  I also have some ROIs as spheres in MNI space images... How can I include it in CONN software ?

  Sincerely,


Gonzalo Rojas Costa

How to restrict my second-level analysis to the connectivity between 2 areas in seed-to-voxel analysis

$
0
0
Hi CONN users!

I want to do a seed-to-voxel analysis between two areas (e.g. dorsal striatum and dorsolateral fronal cortex) but I do not know how to restrict my analysis. I have seen that in the Setup.options there is the possibility to do all the preprocessing just with a mask that you are interested in, but I was wondering if there is an easier and faster way to do this. I also know that through the SPM option in the results window I can put a mask, but this is not what I am looking for. 

Do you know how could I restrict my analysis to see the connectivity between 2 areas using a seed-to-voxel approach? 

Thank you in advance!!

Applying single subject reliability analysis

$
0
0
Dear all 

Is it possible to apply the methods or single subject analysis like in these two papers 

Gordon EM
Laumann TO
Gilmore AW
et al.
Precision functional mapping of individual human brains.
Neuron. 2017; 95: 791-807.e7


using CONN? 


Thanks

Analysis of specific connections

$
0
0
Standard disclaimer: I've searched this forum's archives for any related questions but couldn't find any. My apologies if this has been asked before!

Alfredo et al.: Thanks as always for your continuing work on CONN! My colleagues and I are interested in performing sliding window analyses on data for which we've also obtained structural connectivity. We've found (not surprisingly) that some connections aren't extant in the functional connectivity matrix once conditioned on the thresholded structural connectivity matrix. I don't think CONN has any capability to "mask out" specific connections for analysis, correct? If not, can anyone recommend a way to perform sliding window analyses on (conditioned) ROI x time x subject functional connectivity matrices directly via MATLAB?

And strictly out of curiosity: are there any future plans to include structural connectivity in CONN? Naturally, even the best software can't "do it all," but we've been so pleased with CONN's rigor that we'd be thrilled to see such a feature included.

Regards,
Jeremy Smith.

RE: Extract Connectivity Value Within Cluster

$
0
0
[color=#000000]Hi Steven,[/color]

[color=#000000]From the results-explorer window of your original SBC analysis, simply click on '[b]import values[/b]' and that will compute those average connectivity values per-subject and per-cluster, and it will import them as new second-level variables to your CONN project (if you want, you can then go to [i]Setup.Covariates (2nd-level) [/i]to see those new variables and perhaps use the '[i]Covariate tools. Export selected covariate(s) data to file' [/i]option to export those connectivity values to a file?)[/color]

[color=#000000]Best[/color]
[color=#000000]Alfonso[/color]
[i]Originally posted by Steven Meisler:[/i][quote]Hello,

I have ran a SBC analysis which identified a cluster that I want to analyze further. Specifically, I want to extract the average connectivity values within that cluster for all individual subjects. I have the CON_SubjectXX_compA001_001.nii.gz files from the second-level analyses folder, as well as the significant cluster saved out as a binary mask. However, the dimensions of the cluster mask and CON files are different, so I can't seem to simply use fslmaths to find the average value within that cluster. Does anyone have other suggestions?

Thank you,
Steven[/quote]

RE: gPPI analyses in conn

$
0
0
[color=#000000] Hi Angela,[/color]

[color=#000000]Regarding your first question, yes, in [i]Setup.Conditions [/i]you would simply define your three conditions self/other/case (optionally you may also define the baseline condition), and each condition will encompass data from multiple runs/sessions. Then, when you define your gPPI first-level analysis, simply select the three conditions (self/other/case; there is no need to manually define the regressors that go into gPPI as that will be handled by CONN automatically). If you want to double-check, in the [i]analyses (1st-level) [/i]tab you may switch the 'source timeseries' option to 'first-level analysis design matrix' and that will display the design matrix that the gPPI analysis will use[/color]

And regarding your second question, in option (1) you should upload a [b]single[/b] .txt file, which will include the timing reference for all subjects, sessions and conditions (of course some of these definitions may apply to multiple subjects/sessions/conditions and some will be specific to individual subjects/sessions/conditions; see "help conn_importconditions" for details about the format of this file)

Best
Alfonso
[i]Originally posted by Angela Fang:[/i][quote]Hello,

I have a couple of questions about setting up a gPPI analysis in conn. I have 3 runs of an event-related fMRI task with 3 active conditions (self, other, case) + baseline, per subject. Is it correct that with this design, I would need to include the following regressors: self, other, case, baseline, time series of the seed, PPI_self, PPI_other, PPI_case, and PPI_baseline? When I'm setting up the conditions, would I just set up the self, other, case, and baseline conditions for each subject (with 3 sessions/runs each)? I want to make sure that treating each run as separate sessions for each subject won't create multiple PPI terms and will instead average the data across all 3 runs that contribute to a single PPI term representing all 3 runs. Can you confirm that this is correct? 

Secondly, I'm also wondering how to upload unique timing files for each run in the Setup tab (rather than manually entering all of the onsets and durations for each run for each subject). In the Conditions tab, it looks like there are 3 options for uploading condition info from text files- (1) a Conn-legacy .txt or .csv file, (2) a single BIDS-compatible .tsv file, or (3) a BIDS-compatible .tsv per subject/session. I have a paradigm .txt file that is unique for each run, so would it work to upload the timing file for the entire run even though it includes info about all of the event types- in other words, can I upload a separate .txt file for each subject/run (using option 1), or would I have to create separate event-specific timing files?

Thanks for your help!
Angela[/quote]
Viewing all 6871 articles
Browse latest View live


<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>