Quantcast
Channel: NITRC CONN : functional connectivity toolbox Forum: help
Viewing all 6875 articles
Browse latest View live
↧

Error using cat

March 30, 2017, 2:06 pm
$
0
0
When running thr denoising, at part 4/6, at the 5th participant, the process stops and I get the following message:

ERROR DESCRIPTION:

Error using cat
Dimensions of matrices being concatenated are not consistent.
Error in conn_process (line 1307)
dataroiall{nroi}=cat(1,dataroiall{nroi},y);
Error in conn_process (line 27)
case {'preprocessing_gui','denoising_gui'}, disp(['CONN: RUNNING DENOISING STEP']); conn_process([1.5,2,6:9],varargin{:});
Error in conn (line 3729)
else conn_process('denoising_gui');
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
SPM8 + Beamforming DEM FieldMap MEEGtools
Matlab v.2013b
storage: 354.2Gb available

What can be the source for it and how can it be solved?

Thank you so much!

Maayan
↧
Search

Error with DMN in global source list

March 30, 2017, 3:19 pm
$
0
0
Hello, 

I am attempting to run a seed to voxel analysis with bilateral amygdala seeds, and previously ran a seed to seed approach looking at the default mode network. Now, when trying to run new analyses, I am receiving the error message shown below. I am not sure if CONN is looking for the dmn.nii file I previously used, or if this is some other error, but whenever I attempt to re-run the analysis, I still receive this message:

ERROR DESCRIPTION:

Error using conn (line 5552)
Source DMN.PCC not found in global source list. Please re-run first-level analyses
Error in conn (line 5092)
if any(CONN_x.Analyses(CONN_x.Analysis).type==[2,3]), conn gui_results_s2v;
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 conn dicm2nii wfupickatlas
Matlab v.2016a
storage: 318.8Gb available
Warning: Contents.m overloaded by version in folder C:\Program Files\spm12\toolbox\SPEM_and_DCM


I also tried the fix I saw in a previous post recommended by Alfonso, where you enter the following command in the MATLAB window:
global CONN_x CONN_x.Analyses(1).sourcenames=CONN_x.Analyses(1).source

When I go to my second level results after entering this command, I now can successfully circumnavigate the error message, but it shows the old DMN seeds as my sources rather than the new amygdala seeds which I am now trying to run. Any advice about this is greatly appreciated!

Best, 
Emmaly
↧
↧

FDR p values in ROI-ROI analysis

March 31, 2017, 3:37 am
$
0
0
Dear All,

I would be most grateful if anyone can shed some light on what I expect may be a relatively basic statistical query.

I am performing an ROI-ROI analysis, I have six ROIs that are all entered as both seeds and targets. One of these connections has a uncorrected p-value of 0.0125 and a FDR corrected value of 0.0624.

If I add two more ROIS (so we now have 8 seeds/targets) the uncorrected p-value naturally remains the same. The FDR corrected value however [i]decreases [/i]to 0.0437. Intuitively this does not make sense to me as I imagine that as one increases the number of potential comparisons the FDR p values would tend to get larger.

Any comments appreciated

Best Wishes

Rob
↧

RE: Reporting Results T-Value

March 31, 2017, 8:58 am
$
0
0
Hi, 

Thanks for getting back about this and all of it was completely clear.  So, you wouldn't suggest including the T-values?  When I just quickly looked, I noticed that there were several papers which had also included T-values in their results.    

Also, what is the plot effects table reporting?  I notice that the p values are different from the results window.   

Also, regarding reported the effect size.  Would I still be reporting beta when doing a one-sample T test (for the full group and seed region)?  Just wondering if anyone has any references on reporting effect size or papers that have done this?  

Thanks, 

Jennifer
↧

extract r values roi to roi

March 31, 2017, 1:11 pm
$
0
0
Hi Alfonso,


I am defining a 'reward network' at rest (using coordinates from meta-analysis.) (resting state data)
I have a 2*2*2 ANOVA: one within factor (drug  placebo/drug); 2 between factors (group  depression/healthy) (drug order 1-2/2-1)
I am interested in:
1. Effect of drug
2. Effect of depression
3. Drug*depression interaction
For.....
1. measures of efficiency and centrality within the reward network
2. for hubs (i.e. how connected the ventral striatum is to all the other ROIS in the network, which involves summing the absolute values of all VS-ROI correlations)
 Can I do this in Conn?
How can I extract ROI-ROI correaltion (r values)?

Thank you for your help.
Selina
↧
↧

RE: One structural scan for each functional scan

March 31, 2017, 7:22 pm
$
0
0
Hi Alfonso,

Can you give some instructions about how to change the option in the Setup.functional tab that reads "session-invariant structural data" to "session-specific structural data"? I could not find this option. Thank you so much!

Best,
Lexie
↧

extract denoised time courses

April 2, 2017, 3:21 pm
$
0
0
Dear Alfonso,

I would like to extract the denoised time courses for all regions (AAL atlas and some own ROis). I wanted to use REX, but am confused on what files to use as source and ROI.
Should I use the first-level files "BETA_Subject*_Condition*_Source*.nii as source? What is then the ROI (as the source refers to the ROI).
Or should I use the files in preprocessing: DATA_Subject*_Condition*.mat as source? what would then be the ROI (the ROI_Subject*_Condition*.mat)?

Many thanks
Heidi
↧

RE: Definition on graph theory(statistics)

April 3, 2017, 9:42 am
$
0
0
[color=#000000]Dear Snowa[/color]

[color=#000000]In a two-sample t-test the beta value reported corresponds to the average differences in your measure of interest between the two groups (e.g. average global efficiency within groupA - average global efficiency within groupB). The T- and dof values correspond to the usual T- statistics and degrees of freedom of a two-sample t-test (e.g. dof = number of samples minus 2). The reason why some ROIs may show lower dof values than expected are missing values. In particular average path length and clustering coefficient measures are not well-defined when the node is not fully connected, or when the node has no neighbors, respectively. Those specific cases are treated as missing values and disregarded from the second-level analyses, and that is indicated by reduced degrees of freedom of the corresonding ROI/node (e.g. for a two-sample t-test, dof+2 will be the number of subjects with non-missing data actually included in your second-level analysis for each ROI/node). All other measures (global and local efficiency, centrality, cost, degree) are not affected by this and should show the expected dof values for your test (e.g. N-2 for a two-sample t-test). [/color]
[color=#000000]Last, regarding the edge-width in the picture shown, that plot simply tries to represent the distribution of networks being considered in your analyses (graph-theory second-level analyses will compute a separate network for each subject and for each condition; the properties of these individual networks are the ones that are being tested by your second-level analysis model). In particular the thickness represents the proportion of cases where that edge is present (among all subjects/conditions), and only edges with above 25% values are displayed.  Note that this is just for plotting purposes, so the edge widths are not really related to the actual models being tested, they are there just to give you an idea of how fully- or sparsely connected are the networks that you are looking at. On the other hand node sizes are in fact related to T- statistics associated with the selected measure for that node (e.g. for a two-sample t-test on global efficiency, nodes that show stronger/more-significant between-group differences will be displayed with larger sizes)[/color]

[color=#000000]Hope this helps[/color]
[color=#000000]Alfonso[/color]
[i]Originally posted by snowa wo:[/i][quote]Dear Alfonso.
(I do not have an answer, so I write again..sorry)
 
When doing two-sample t-test in graph thoery,
statistical values were delicered such as beta, T and dof.
What beta and dof does it mean?
i know the beta value is the value from the regression analysis.
i do not know why the beta value appears in the t-test.
And i do not know why dof values comes out differently in each ROI (not whole network)
When I ran the t-test with 52 samples(group A=25, group B=27),
The dof value of the whole network is 50.
However, the dof value of ROI in the row below are 42, 46.. and so on.
Are ROI below the whole network not the result of t-test?
Additionally, what determines the node size and edge thickness in 3D image of graph theory?
(ex) edge thickness of 3D image reflects T value)
 
thank you!
 [/quote]
↧
Search

RE: Seed to voxel analysis problem

April 3, 2017, 9:49 am
$
0
0
[color=#000000]Hi Rob,[/color]
[color=#000000]Not really enough information to make a guess. If you want to send me your two conn_*.mat files I will take a quick look to see if I can figure out the source of discrepancy. [/color]
[color=#000000]Best[/color]
[color=#000000]Alfonso[/color]
[i]Originally posted by Rob McCutcheon:[/i][quote]I have ran a resting state analysis twice (once on a network and once on a local machine)

When looking at the 1st level results, and the 2nd level ROI-ROI results, the findings are very similar, however there are marked discrepancies in the seed to voxel analyses.

In analysis A the results appear normal, however in analysis B almost nothing shows up on the seed to voxel analyses. When one plots values, the beta maps look normal for analysis A, but they are barely visible in analysis B. In both cases the threshold is the same (P uncorrected <0.001).

The only difference i can think of is that for analysis B I preprocessed in CONN whereas in analysis A i did it separately in SPM. I think this is unlikely to be the cause though given that the 1st level results are very similar for both analyses.

I wondered if anyone has encountered this before or might be able to shed light on where I have made an error?

Many thanks

Rob[/quote]
↧
↧

RE: Error using cat

April 3, 2017, 10:07 am
$
0
0
[color=#000000]Hi Maayan,[/color]

Not totally sure, could you check:

1) that the functional data entered for this subject looks fine (e.g. this error may arise if your BOLD signal timeseries is exceedingly short, e.g. less than 16 timepoints). To do this, for example go to [i]Setup.Functional[/i], click on this subject sessions, and look above the brain display image the "number of files" reported.

and 2) that the White/CSF masks for this subject look fine (e.g. this error may also appear, I believe, if your eroded white matter or CSF masks are almost but not totally empty, which could indicate just a suboptimal normalization of this subject's structural data). To do this, for example (this requires 17c) go to [i]Setup.ROIs, [/i]select CSF, click on 'erosion settings' and 'Ok' and that will re-compute the eroded masks for your subjects and it will report the number of suprathreshold voxels in these masks. 

Thanks 
Alfonso
[quote]When running thr denoising, at part 4/6, at the 5th participant, the process stops and I get the following message:

ERROR DESCRIPTION:

Error using cat
Dimensions of matrices being concatenated are not consistent.
Error in conn_process (line 1307)
dataroiall{nroi}=cat(1,dataroiall{nroi},y);
Error in conn_process (line 27)
case {'preprocessing_gui','denoising_gui'}, disp(['CONN: RUNNING DENOISING STEP']); conn_process([1.5,2,6:9],varargin{:});
Error in conn (line 3729)
else conn_process('denoising_gui');
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
SPM8 + Beamforming DEM FieldMap MEEGtools
Matlab v.2013b
storage: 354.2Gb available

What can be the source for it and how can it be solved?

Thank you so much!

Maayan[/quote]
↧

RE: Second level analysis error

April 3, 2017, 11:56 am
$
0
0
[color=#000000]Dear Larissa,[/color]

[color=#000000]Regarding (1) t[/color]hanks for reporting this bug, please try the attached patch and let me know if you still run into any issues (this patch is for release 17c, simply copy this file to the conn distribution folder overwriting the file with the same name there)

Regarding (2) when selecting all four seeds CONN is performing a multivariate test (instead of a univariate test for a single seed) looking at any effect across the four seeds. The multivariate test is going to automatically correct for the "multiple-test" scenario of having four seeds. It is mathematically possible but in practice somewhat unlikely that you will get a significant multivariate effect in the absence of any univariate effect. Could you confirm if, when looking at the univariate effects (separately for each seed), you are selecting a "two-sided" test? (by default the results explorer window for a univariate test comes up with the one-sided test -positive directionality- just to follow SPM convention, but in this case I am thinking that perhaps you may not be seeing any univariate effect because they may be negative / in the opposite directionality?)

[color=#000000]Hope this helps[/color]
[color=#000000]Alfonso[/color]
[i]Originally posted by Larissa :[/i][quote]ERROR DESCRIPTION:
Dear Alfonso,

I have 2 questions:
1. When running a seed-to-voxel analysis comparing 2 resting-state conditions in 10 subjects I get the following error message when selecting 'Results for all sources.'  Note I'm doing an exploratory analysis: my between subjects contrast is 1 (for All subjects), Between conditions contrast is [1 -1] when comparing the 2 conditions. 

Index exceeds matrix dimensions.
Error in conn_process (line 3937)
nrepeated=size(SPMall(n1).xY.VY,2);
Error in conn_process (line 51)
case 'results_voxel', [varargout{1:nargout}]=conn_process(16,varargin{:});
Error in conn (line 6681)
conn_process('results_voxel','doall','seed-to-voxel');
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.2012a
storage: 1530.9Gb available

2. If I want to look at a specific network as an a priori seed/source for the seed-to-voxel design (e.g. the visual cortex), there are 4 network ROIs that CONN provides (ie. network.Visual.Primary/Ventral/Dorsal (L)/Dorsal (R). If I select all 4 visual networks as the seed ROI, I get a significant cluster as a result when selecting Results explorer. However, when I select each individual network ROI on its own, I do not find any significant clusters. My understanding is that my ROI is larger and is therefore more prone to find significance when selecting all 4 ROIs that contribute to the network in question. Does CONN do a multiple comparisons analysis with FDR if this is the case? Is it normally justified to analyze data as such, since the 4 ROIs contribute to the same network?

Thank you,
Larissa[/quote]
↧

Preprocessing Centering

April 4, 2017, 3:45 pm
$
0
0
Hello, 

I am running the preprocessing steps in CONN and I notice that the default preprocessing includes a centering option...

"functional Center to (0,0,0) coordinates" and "structural Center to (0,0,0) coordinates"

Is this step required if I've already manually put my images to the AC and if I include centering in my preprocessing steps, will it mess up the preprocessing?  

Thanks again,

Jen
↧

RE: New subjects create NaN 2nd level covariates

April 4, 2017, 7:27 pm
$
0
0
[color=#000000]Thanks Alfonso, that fixed my issues; Appreciate your quick and clear response![/color]

[color=#000000]- Matt V.[/color]
[i]Originally posted by Alfonso Nieto-Castanon:[/i][quote][color=#000000]Hi Matt,[/color]
[color=#000000]Sorry about this, I believe that issue had been reported and fixed in release 17b. If you can update to the latest release and repeat the analyses I would recommend that, otherwise let me know which CONN version you have and i will send you instructions on how to reconstruct the replaced-by-NaN values for those covariates.[/color]
[color=#000000]Best[/color]
[color=#000000]Alfonso[/color]

[i]Originally posted by Matt Vandermeer:[/i][quote]Hi all,

I'm relatively new to using CONN and fMRI data-analysis en masse. That said, I have been conducting preliminary analyses and everything was going fairly smoothly. I recent updated my set-up stage to include newly collected data from new research subjects; however, following pre-processing and denoising, the QA_MaxMotion, QA_InvalidScans, and QA_ValidScans second-level covariate values for all previously analyzed subjects has been replaced with "NaN" values. Can anyone explain this to me and hopefully point me in the right direction for finding these previously available values? Thanks![/quote][/quote]
↧
↧

plot setting in graph theory

April 5, 2017, 12:43 am
$
0
0
Dear Alfonso,
 
I am setting threshold empth in the gui (on graph thoery) and pressing enter,
and wanted to global and local efficiency line of the networks resulting from "each group".
In other words, I want to see the LINE of each group's results in ONE plot.
However, despite the fact that set the t-test, in the graph, my groups are displayed as a line with the name "data".
Which option should I change?
 
Best,
Snowa.
↧

seed areas for resting state networks

April 5, 2017, 3:54 am
$
0
0
Dear all,
 
can somebody tell me how the seed areas in CONN for the investigation of resting state Networks were generated?
Did they use 3T MRI data?
 
Thank you in advance!
↧
Search

Setup Pipeline error: "mismatched dimensions"

April 5, 2017, 5:42 am
$
0
0
Hello,

I preprocessed my data with conn successfully. Then I wanted to proceed to the "Setup" step: I clicked "Done", then tried to run the default Setup pipeline, upon which I received the following two error messages for every subject:[quote]ERROR: Subject XXX Session 1 first-level covariate realignment mismatched dimensions (250 rows, while functional data has 245 scans; the number of rows of a first-level covariate should equal the number of scans for this subject/session)

ERROR: Subject XXX Session 1 first-level covariate scrubbing mismatched dimensions (250 rows, while functional data has 245 scans; the number of rows of a first-level covariate should equal the number of scans for this subject/session)[/quote]
This is my preprocessing pipeline (in order):[quote]-functional Slice-time correction
-functional Realignment&Unwarp
-functional Center to (0,0,0) coordinates
-structural Center to (0,0,0) coordinates
-functional coregistration to structural
-structural segmentation&normalization
-functional normalization
-functional outlier detection (ART-based scrubbing)
-functional smoothing
-functional removal of initial scans (first 5 scans)[/quote]
Due to the apparent lack of 5 scans I thought the problem might be due to my last preprocessing step (removal of first 5 functional scans). But then why would conn offer this preprocessing option if it were not able to further process the resulting data?

I am using CONN v.16.b

What is the problem here?

Thanks!

Best,
Sascha
↧

connectome-MVPA reference

April 5, 2017, 6:03 am
$
0
0
Hi everyone and hi Alfonso,

I was wondering if there is a reference that I can use when describing the connectome-MVPA measure in a manuscript

Alain
↧
↧

RE: realignment&unwarp

April 5, 2017, 6:37 am
$
0
0
Dear Dr. Alfonso,

thank you!

Best,
Sascha
↧

RE: Extract first principal component only (PCA)

April 5, 2017, 7:30 am
$
0
0
Dear Conn-Users

I have a short question. I am writing a batch script and I would like to use the first principal component of my ROI only.

In the conn_batch_help document it says:
dimensions : rois.dimensions{nroi} number of ROI dimensions - # temporal components to extract from ROI [1] (set to 1 to extract the average timeseries within ROI voxels; set to a number greater than 1 to extract additional PCA timeseries within ROI voxels; set to 0 to compute a weighted sum within ROI voxels (ROI mask values are interpreted as weights))
From the description it is unclear to me, to which number I need to set it to only get the first principal component? 

A short advice would be greatly appreciated!

Isabel
↧

Subject space analysis - normalise beta maps?

April 6, 2017, 3:39 am
$
0
0
Dear Alfonso and others,

I am attempting a resting state analysis in which I wish to perform the first level analysis in subject space (I have previously performed it in MNI space, and am wanting to check that the results remain the same with this method).

In my preprocessing I center the images, realign and unwarp the functional, slice time correct, coregister, segment the structural, and use ART for outlier detection. I then use the i_y*.nii files generated during the segmentation to move my ROIs (originally in MNI) to subject space, which I do for each subject

My questions regards how to perform the second level analysis - how do I normalise/smooth the beta maps to MNI space for this step? If I just go to the second level analysis tab as things stand my results currently do not show the connectivity patterns I expect (e.g. just a few random scattered voxels for the second level results).

Many thanks,

Rob
↧
Viewing all 6875 articles
Browse latest View live
Search