Hi Dr. Spielberg,

I've encounterted an error during the preprocessing steps and was wondering whether you may be able to help me with it. Below is the error in full:

Warning: matlabpool will be removed in a future release.
Use parpool instead.
Starting matlabpool using the 'local' profile ... connected to 2 workers.
Undefined function or variable "WM_base_filename".

Error in GTG_preprocess_GUI>Start_pushbutton_Callback (line
1276)
parfor currsub = 1:length(subs)
Error in gui_mainfcn (line 95)
feval(varargin{:});
Error in GTG_preprocess_GUI (line 134)
gui_mainfcn(gui_State, varargin{:});
Error in
@(hObject,eventdata)GTG_preprocess_GUI('Start_pushbutton_Callback',hObject,eventdata,guidata(hObject))
Error while evaluating uicontrol Callback

Could you please let me know which may have caused this?

Also, this may or may not be related to the error: my 3D ROI mask contains 6 nodes, which is indicated in the cell array of ROI labels. In the documentation, it says "Each ROI must have a unique numeric identifier (identifiers need not be consecutive) and must match the #'s entered in the array of ROI labels entered above". Does the "identifiers' refer to the numeric values in the second column of the ROi label cell array or does it refer to something else?

Thank you for your help!

A.

Hi Alex,

Are you trying to partial signal from white matter?  If so, did you specify the filename for the white matter mask for first participant?  

RE your second question, yes the 'numeric identifiers' are the numbers with which the different ROIs are labelled in your ROI mask, which corresponds to the second column in the ROI label cell array.  

Best,
Jeff

Thank you for your response, Dr. Spielberg. I am not partialing out signal from white matter. The option was left unchecked in the GUI and the file selection was greyed out.

W/r/t to my second question, how does GTG know which ROI corresponds to which numeric value in the actual mask? I have 6 separate ROIs which have been combined into a single mask. This is the file I have selected for the 3D ROI mask. Sorry if this is a naive question.

A.

Ok, I'll try to duplicate the error.  What options you did check (you can just send a screenshot of the GUI with options checked)?

RE your second question, each ROI in your mask should consist of a different number.  So, for example, let's say that you have (right and left) amygdala, hippocampus, and caudate.  Then, your mask could be something like this: voxels in right amygdala have a value of 1, voxels in left amygdala a value of 2, 3 for right hippocampus, 4 for left hippocampus, 5 for right caudate, 6 for left caudate.  Then, your labels cell array would look something like this:

Ramyg  1
Lamyg  2
Rhipp    3
Lhipp    4
Rcaud   5
Lcaud   6

The script would then know that voxels in your mask that contain a value of 1 correspond to right amygdala, etc.

I have attached here a screenshot of the GUI.

And thank you for the explanation on the ROIs. That clears things up.

A.

To make a long story short, matlab has some non-logical requirements for parallel for loops, and the script didn't meet of these.  I'll put a fix in the next version, but for now try replacing with the attached script and let me know if this fixes your problem.

Best,
Jeff