using a mask in conn/utils/otherrois/buildP.m

December 18, 2017, 5:07 am

≫ Next: RE: Single subject analysis

≪ Previous: why are my images displayed very small

$

0

0

Dear Alfonso,

I am trying to use the buildP.m provided in CONN/utils/otherrois.

My idea is to generate 1.5mm cubes only within a mask (e.g. within DMN or other networks) instead of the whole brain. I tried firstly to generate the 1.5mm (voxel size) whole brain parcels and then mask them with DMN. However, If I set the sidelengths to 1.5 mm in buildP.m, my Matlab crashed (maybe due to too many computations).

I also tried to set the "bounding box" (bbox) in buildP.m to the range of each of my ROI, but the output was wrong (out of the range of my ROI).

So the question is [b]how to overlap a mask during the process of buildP.m[/b].

Thanks a lot for your help!

Best regards,
Yifei

---

RE: Single subject analysis

December 18, 2017, 5:38 am

≫ Next: RE: REX Results GUI: Meaning of Effect Sizes

≪ Previous: using a mask in conn/utils/otherrois/buildP.m

$

0

0

Dear Steven, 

According to what Alfonso wrote, :"The files named "BETA_Subject#_Condition#_Source#.nii" (...) contain the Fisher-transformed correlation values between each seed/source and the rest of the brain (for each subject, condition, and seed/source).". I was wondering why would you suggest thresholding the results on Fisher transformed r of 0.25? Shouldn't we threshold it by the p-value (FDR corrected) to take under account the significance of correlation in each voxel?

Thanks! Kasia

RE: REX Results GUI: Meaning of Effect Sizes

December 18, 2017, 6:12 am

≫ Next: Compatibility CONN toolbox with Matlab R2017b

≪ Previous: RE: Single subject analysis

$

0

0

Dear Alfonso 

I have a follow up question, but first I want to thank you again for your previous help. We ran the analyses as you recommended. Now a reviewer pointed out that our analysis as above indicates that our treatment effect are relative changes in connectivity and symptom scores. He/she asked us to give more explanations form a mathematical perspective about fisher transformation correlation values and statistical results.

Where do I find such values?

Kind Regards
Julian

Compatibility CONN toolbox with Matlab R2017b

December 18, 2017, 9:54 am

≫ Next: missing BATCH.Setup.preprocessing.sliceorder in conn_batch_humanconnectomeproject.m?

≪ Previous: RE: REX Results GUI: Meaning of Effect Sizes

$

0

0

Dear community,

as I my mac was recently uploaded to High Sierra (version 10.13.2) I will soon have to use Matlab R2017b which apparently works most reliably with this macOS version. Does anybody of you have already made experience whether the COON toolbox works with the mentioned versions of macOS and Matlab?

Thanks your help, kind regards,
J. Rosen

missing BATCH.Setup.preprocessing.sliceorder in conn_batch_humanconnectomeproject.m?

December 18, 2017, 10:23 am

≫ Next: Outputs change folders

≪ Previous: Compatibility CONN toolbox with Matlab R2017b

$

0

0

Hello,

Concerning the sample script "conn_batch_humanconnectomeproject.m", is it missing the line that defines slice order?

Based on the Human Connectome Project Sessions Protocol (https://humanconnectome.org/storage/app/media/documentation/s1200/HCP_S1200_Release_Appendix_I.pdf page 40), should the code have the following line?[code]batch.Setup.preprocessing.sliceorder = 'interleaved (Siemens)'[/code]
Cheers,
Walter

Outputs change folders

December 18, 2017, 12:19 pm

≫ Next: Connectivity matrix for task data

≪ Previous: missing BATCH.Setup.preprocessing.sliceorder in conn_batch_humanconnectomeproject.m?

$

0

0

Regarding outputs, is there a way to send the output images (realigned, normalised, etc..) directly to a different folder from the original ones? This is because I want to preprocess the same subjects using different QA thresholds and then compare them. Otherwise I think I just could replicate the original images in different folders and start each time from one of these folders.

Note: I do not know if this question should correspond to a new post or not.
Note 2: As a relateively new user in conn I am impressed of how active and helpful this forum is. Thanks developers and users!

Connectivity matrix for task data

December 19, 2017, 4:25 am

≫ Next: Information on ROIs provided by CONN

≪ Previous: Outputs change folders

$

0

0

Dear CONN experts,

I have a question regarding the calculation of connecitivity matrices in CONN. I am analysing task data with two conditions: during one scanning session, participants experience periods of a task and rest. I understand that I can find the connectivity matrices for each participant and each condition in the results/firstlevel folder.

Now my question: how are these connectivity matrices calculated for the different conditions? For example, how is the time of task onset determined - is the HRF taken into consideration for this? Are the 1st level covariates taken into consideration when calculating the matrices? For example, is the correlation matrix calculated using the BOLD signal with motion signals regressed out? 

I apologise if this is a very basic question, however I was looking through the documentation and forum and unfortunately I could not find an answer.

Best
Judi

Information on ROIs provided by CONN

December 20, 2017, 3:01 am

≫ Next: Impact of sphere size (4 vs. 6 or 8 or 10 mm)

≪ Previous: Connectivity matrix for task data

$

0

0

Hi, 

I have some questions about the atlas and networks ROIs provided by CONN:

Is it possible to confirm that when using these ROIs:

-Networks ROIS: the average signal from a 10 mm sphere around the provided coordinates

-Atlas ROIS: the average signal from all voxels in the whole anatomical region (and same for AAL or BA atlases?)

Thank you so much!

---

Impact of sphere size (4 vs. 6 or 8 or 10 mm)

December 20, 2017, 3:04 am

≫ Next: Preprocessed and denoised data

≪ Previous: Information on ROIs provided by CONN

$

0

0

Hi CONN users,

can anyone help me figuring out the impact of sphere size on the results? What are the advantages/disadvantages of using smaller vs. larger spheres? 

Anything would help (experiences, references, etc.) 

Thank you very much!

Preprocessed and denoised data

December 20, 2017, 6:45 am

≫ Next: RE: Changing the atlas and MNI positions

≪ Previous: Impact of sphere size (4 vs. 6 or 8 or 10 mm)

$

0

0

Dear Alfonso
Dear All

I wonder where I can find in Conn the preprocessed, denoised, band-filtered and possibly z-scored subject/session data? And in which format? For further use in other software.
I would be grateful for your answer
Lucas

RE: Changing the atlas and MNI positions

December 21, 2017, 1:27 am

≫ Next: REX

≪ Previous: Preprocessed and denoised data

$

0

0

[color=#000000]Dear All[/color]

Where can I find the MNI coordinates correspond to the predefined regions listed in the atlas.txt?

Thanks for considering my request.


Best wishes
Larry

 [i]Originally posted by Alfonso Nieto-Castanon:[/i][quote][color=#000000]Dear Farras,[/color]

[color=#000000]To use a different ROI atlas file you may simply select the new atlas file from the gui (e.g. in Setup->rois, simply enter a new ROI line in the ROIs list, enter a name for this atlas, and then select all subjects and select the atlas116.img file) or use a batch script entering the new atlas information in the field batch.setup.rois (see the batch manual for more details). Yet another option is to simply include the atlas116.img file in the conn/rois/ folder and that will include all the ROIs in this atlas by default on every new project that you create. Any of these options will use the ROIs defined in your atlas file as new ROIs (for seed-to-voxel and/or ROI-to-ROI analyses). The line that you are changing, nevertheless, is used for display/information-purposes only (when you right-click on a voxel in any whole-brain display within the conn gui, it will tell you its spatial coordinates as well as the BA area that this voxel may belong to). Unfortunately I have not included yet a simpler option to change the default BA atlas used for these displays, so if this is what you would like to change, I am afraid that for now editing that line is the simplest option to achieve this (I will make sure to add in the next release of the toolbox a simpler option to allow you to use other atlases for these displays). [/color]Let me know if this clarifies.

[color=#000000]Best[/color]
[color=#000000]Alfonso[/color]

[i]Originally posted by Farras Abdelnour:[/i][quote]Dear all,
I've been using an atlas with 116 ROI on conjunction with conn.m. For this, I needed to slightly modify the program around line 267 as follows:

filename = fullfile(fileparts(which('conn')),'rois',''atlas116.img');

As I will be using different atlases, is there a way to feed conn.m the atlas type without having to actually modify the program every time a new atlas is used?

With thanks,

Farras[/quote][/quote]

REX

December 21, 2017, 7:27 am

≫ Next: Slice timing 'Manually define'

≪ Previous: RE: Changing the atlas and MNI positions

$

0

0

Hello Dears,
I have used REX to generate the connectivity values for each ROI and the output files for each ROI is a set of numbers (as stated below). I was wondering if someone could tell me what each value is indicating? Thank you guys in advance!

0.4950
0.5442
0.5614
0.5679
0.5952
0.5738
0.5604
0.5041
0.5361
0.4983
0.4703
0.4881
0.4113
0.4184
0.4848

Slice timing 'Manually define'

December 21, 2017, 8:02 pm

≫ Next: Cluster extent thresholding of first-level connectivity maps

≪ Previous: REX

$

0

0

Dear CONN-experts

During slice time correction, let's say that my slice order is in a default setting on Philips Intera 3T scanner (48 slices), so 1,3,5,7,...2,4,6,8 etc.

When using the GUI, in the slice-timing correction prompt select 'Manually define' and automatically returned as 1,2,3.......91.

Number of slices is... : 91
Time to Repeat (TR) is... : 3
Parameters are specified as... : slice indices
Completed : 11:46:44 - 22/12/2017
Done 'Slice Timing'
Running 'Slice Timing'

I cannot proceed the slice-timing with [1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48].

Any suggestion?

Thanks for considering my request.

Cheers
Larry

--
Philips Interga 3T (with default scan order):

Repetition time (ms): 3000
Echo time[1] (ms): 30.001
Flip angle: 80
Number of averages: 1
Slice thickness (mm): 3.313
Slice spacing (mm): 3.313
Image columns: 64
Image rows: 64
Phase encoding direction: COL
Voxel size x (mm): 3.3125
Voxel size y (mm): 3.3125
Number of volumes: 140
Number of slices: 48
Number of files: 6720
Number of frames: 0
Slice duration (ms) : 0
Volume interval (s): 46
Orientation: tra

Cluster extent thresholding of first-level connectivity maps

December 22, 2017, 3:55 am

≫ Next: error in pre-processing step after flipping left/right hemispheres

≪ Previous: Slice timing 'Manually define'

$

0

0

Dear CONN community, 

In my single case study I would like to report first level maps of my patient. I have already figured out (with a big help from this forum, for which I am very grateful!) how to extract first level maps with correlation values and how to threshold them with their corresponding voxel-wise p values, uncorrected and FDR corrected. 

However now I would also like to be able to apply a cluster-wise threshold, either with the FDR correction or just simply with a cluster size value. I wasn't able to figure it our by myself, has anyone done that already? I would be very grateful for your help. 

All the best, Kasia

error in pre-processing step after flipping left/right hemispheres

December 25, 2017, 8:16 pm

≫ Next: CONN's non-parametric and TFCE toolbox

≪ Previous: Cluster extent thresholding of first-level connectivity maps

$

0

0

Hello everyone,
I am facing the following error in the pre-processing step in CONN version 17.f. I would appreciate if you let me know why is that and how I can fix it.
a background about what I have done is:  I have 9 subjects , with pre/post sessions.  one of the subjects has only pre, and doesn't have post. so I checked the "allow missing data" box. In the preprocessing first I flipped left and right hemispheres for functional and structural images by choosing none rigid reflection along x-axis, and it worked with no issues and flipped the images.
After that, I went through preprocessing again, and this time I selected default pre-processing, and after a while it gives me the following error. I also got many warnings in the Matlab Command Window including: "Warning: Matrix is singular to working precision". , and " Warning: Forcing deletion of MAT-file"


ERROR DESCRIPTION:

Error using spm_jobman>fill_run_job (line 461)
No executable modules, but still unresolved dependencies or incomplete module inputs.
Error in spm_jobman (line 247)
sts = fill_run_job('run', cjob, varargin{3:end});
Error in conn_setup_preproc (line 1940)
job_id=spm_jobman('run',matlabbatch);
Error in conn (line 894)
ok=conn_setup_preproc('',varargin{2:end});
Error in conn_menumanager (line 120)
feval(CONN_MM.MENU{n0}.callback{n1}{1},CONN_MM.MENU{n0}.callback{n1}{2:end});
CONN v.17.f
SPM12 + DEM FieldMap MEEGtools
Matlab v.2016b
storage: 1116.6Gb available
spm @ E:\NEUROSCIENCE\spm12\conn17f\conn
conn @ E:\NEUROSCIENCE\spm12\conn17f\conn


also, the last line in the Matlab command when I get the above error is:  "Item 'Session', field 'val': Number of matching files (1) less than required (2)."
P.S. I am not getting error when I simply use the pre-processing, without first flipping left and right hemispheres.

Any help will be highly appreciated.

---

CONN's non-parametric and TFCE toolbox

December 29, 2017, 5:34 pm

≫ Next: Help with

≪ Previous: error in pre-processing step after flipping left/right hemispheres

$

0

0

[color=#000000]Dear CONN-experts[/color]

This is what I have done so far.

[list=1][*]Original F-test: eye(5)-1/n for Group1, Group2, Group3, Group4, Group5 (Looking for effects of interest)[*]Between group difference: [1 -1 0 0 0; 0 1 -1 0 0; 0 0 1 -1 0; 0 0 0 1 -1] (step (1) & (2) yield similar results)[*]Performing kruskal-wallis test to compare the behavior score based on demographic characteristics.[*]Post hoc two-sample t for step (3) with significant clusters found in step (2), and then performing CONN's non-parametric or TFCE[*]Here comes a question, does it make sense to perform CONN's non-parametric or TFCE toolbox for step (1) and (2)? or should I stick it to step (4) to avoid multiple testing issues?[/list]
Thanks for considering my request and questions.

Have a lovely weekend and happy new year!

Best wishes
Larry
[i]Originally posted by Alfonso Nieto-Castanon:[/i][quote][color=#000000]Hi Ely,[/color]

[color=#000000]Just for reference, if the main concern driving the suggestion to use TFCE is to avoid parametric assumptions about cluster-size distributions then you may also consider using CONN's non-parametric analyses for that (these use permutation/randomization tests and are available in the latest CONN release; see top-right menu in the seed-to-voxel results explorer) and use the associated cluster-size or cluster-mass thresholding options there.[/color]

[color=#000000]Best[/color]
[color=#000000]Alfonso[/color]
[i]Originally posted by Benjamin Ely:[/i][quote]You can disregard this post - I think the issue I was experiencing had to do with some changes I made to the file tree. When I started the analysis over from the beginning, the first-level BETA maps for the mulitvariate and bivariate analyses were different, as expected. Entering these first-level multivariate maps into FSL's randomise function then allowed me to perform TFCE.

Thanks anyway, though!
-Ely[/quote][/quote]

Help with

January 2, 2018, 6:44 am

≫ Next: How to use mask on CONN

≪ Previous: CONN's non-parametric and TFCE toolbox

$

0

0

Hi all, 

I have been using to make comparisons between two groups of patients based on cognition state. In the second level analysis, I make a [1 -1] contrasts and then check the differences using both ROI-to-ROI and Graph Theory analysis. I am getting interesting results. However, I would like to tests other measures such as modularity, disconnection, in these two groups and in different subgroups within my sample. 

My question is: does conn generate any file that contains separate correlations matrices of my two groups? For instance, GAT accepts a resultsROI_condition.mat file, but I can only find one such file in my firstlevel folder, and therefore I guess it contains data for the whole sample. 

Is there any easier way of doing this that I am missing?

Thanks for your time!

IA.

How to use mask on CONN

January 3, 2018, 7:17 am

≫ Next: CONN analysis: group comparision after intervention

≪ Previous: Help with

$

0

0

Hi,

I work with stroke patients and I dont know in which step should I insert the lesion masks on conn. Someone could help with this information? 

- How can I work with lesion masks on conn and how would be this process?

CONN analysis: group comparision after intervention

January 4, 2018, 7:58 am

≫ Next: Running mediation with CONN

≪ Previous: How to use mask on CONN

$

0

0

Dear experts,

The analysis I am trying to perform should show whether a certain intervention shows an effect. We have 2 groups who we want to assume are equal at time point 1. Each group is assigned to an intervention, let's call it "placebo" and "agent". Now I would like to determine whether at time point 2 the "agent" group indeed does show a different connectivity pattern compared to the "placebo" group. Can this be done in CONN, how would I enter this in the second-level/ how would the between-subject contrast and between-conditions contrast need to be defined?

What I have done so far is I have uploaded 2 sessions (time point 1 and 2) for my functional data for every subject, added the functional ROIs I am interested in, entered 2 conditions during the set up step, Tp1 and Tp2 and allocated the subjects according to their group in second-level covariates.

Now how can I determine whether there were changes caused by the intervention, since I need to make sure that any group differences at time point 1 are not accounted for, but only those that appeared after the intervention at Tp2?

I would be immensely grateful for some advise.

Thank you very much!
Meike

Running mediation with CONN

January 5, 2018, 6:48 am

≫ Next: ART-based scrubbing procedure

≪ Previous: CONN analysis: group comparision after intervention

$

0

0

Hi Alfonso,

I have done all the preprocessing and ROi-to-ROI and seed-based analyses of my data.
Now I would like to investigate whether the functional connectivity between two regions is mediated by pathology (from a questionnaire).
What is the best way to approach this? I can do the direct connection between region A and B, but how to enter the scores is a bit tricky.
Basically the model would be:
Region A --> pathology --> Region B
Many thanks!
All the best
Heidi