Run Control

The process of analysis is controlled by a single ExSetup file, where all the settings and definitions are located. Running ExSetup will load all the parameters and eventually call the sc_initAnalysis function, which controls all analytic stages.

Sample EXSetup files can be found in the ExSetupTemplates subdirectory. They contain most of the required option settings with short descriptions. You can easily adapt a sample file for your own analysis.

ExSetup can be divided into several sections. The first one includes parameters grouped as Run Control (rc) which define main parameters, like file locations, stages and datasets to be included

The whole analysis is divided into the following steps:

0: raw file read and preprocessing;

1: segmentation;

2: sensor space artifact rejection;

3: ICA1 decomposition;

4: ICA1 artifact rejection;

5: ICA2 'clean' decomposition;

6: IC classification;

7: realignment, head model, source model;

8: leadfield and localization;

9: ROI signals reconstruction;

10: connectivity estimation;

11: statistics, visualization, data export;

Particular stages of analysis to be run are selected by setting the parameters:

% analytic steps to run
rc.load = 0;  % start from stage
rc.stop = 5;  % stop at stage  

Typically, all datasets present in relevant directories are processed. To limit this range (for example, for testing purposes or setting up a new analysis) , the following parameters are set up:

% keep begId and endId commented, unless you want to analyze only subset of datasets
begId = 1;  %specify the first dataset index to start the analysis from
endId = 6;  %specify the last dataset index to stop the analysis on

To process all available datasets, these parameters should be commented out.

Run control settings include the main directory of the analysis: rc.expFolder. All other subdirectories are then relative to this main directory. Importantly, rc.analysisName defines a separate analysis that will be saved in a subfolder of the same name.

Within the main experimental folder, multiple analysis versions can be performed. This gives an opportunity for multiverse analyses run on the same data or studying different aspects of the same recording (for example, stimuli and response analyzed separately).

This section also includes MRI parameters, which are optional and valid only when individual T1 MRI scans are to be used for source reconstruction.

%% FILE LOCATION
rc.expFolder = '/mnt/iSCSI1/Exp/Flanker';   % main exp folder. everything will be saved within this direstory. all path settings are relative to this directory
rc.analysisName='ana1';                     % the name of the current analysis. everything will be saved with a folder with this name
rc.dataFormat = 'bdf';                      % raw data format: 'itab' 'ctf' (MEG) 'bdf' 'eeglab' (EEG) 'itab-sim' (simulations)
rc.dataRawFolder = '0_raw';                 % raw data folder
rc.mriRawFolder = '0_mri';                  % location of T1 scans. leave empty if average brain template is to be used
rc.mriExt = 'nii';                          % format of T1 scans
rc.mriCoordsys = 'ras';                     % coordsys of MRI scans. typically 'ras'

Parallel computing can significantly speed up some stages of data processing at the cost of increased memory demand. It is enabled with rc.useParallel. Set the number of parallel threads to launch or set to zero to disable.