User Tools

Site Tools


mdm.computeglm

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision
Previous revision
Last revisionBoth sides next revision
mdm.computeglm [2010/06/09 16:20] – created jochenmdm.computeglm [2010/11/24 05:13] – updated the help jochen
Line 2: Line 2:
  
 ===== Motivation ===== ===== Motivation =====
-To combine several subjects' data into one compound analysis, it is custom to compute random-effects statistics on the second level. The MDM file format of BrainVoyager QX references the required time-course files (VTC or MTC) and the respective information containing the model (onsets in PRT files or regressors in SDM files).+To combine several subjects' data into one compound analysis, it is custom to compute random-effects statistics on the second level. The MDM file format of BrainVoyager QX references the required time-course files (VTC or MTC) and the respective information containing the model (onsets in [[xff - PRT format|PRT]] files or regressors in [[xff - SDM format|SDM]] files).
  
 This method then allows to perform the regression on a session-by-session basis and combine the results in one compound GLM object. This method then allows to perform the regression on a session-by-session basis and combine the results in one compound GLM object.
Line 9: Line 9:
 This function is currently implemented for BrainVoyager QX's GLM format only. The following files must be available: This function is currently implemented for BrainVoyager QX's GLM format only. The following files must be available:
   * VTC (or MTC) files for all runs of all subjects (e.g. the ones automatically created by spm5_preprojobs, for how to name files see below)   * VTC (or MTC) files for all runs of all subjects (e.g. the ones automatically created by spm5_preprojobs, for how to name files see below)
-  * PRT (or the derived SDM) files for all runs (if subjects share this information, only one file is needed for each uniquely specified model!) +  * [[xff - PRT format|PRT]] (or the derived [[xff - SDM format|SDM]]) files for all runs (if subjects share this information, only one file is needed for each uniquely specified model!) 
-  * optionally the realignment parameter files (supported formats: TXT for SPM rp*.txt files, SDM for BV's motion correction output)+  * optionally the realignment parameter files (supported formats: TXT for SPM rp*.txt files, [[xff - SDM format|SDM]] for BV's motion correction output)
  
 **Note: The combination of within-subjects regressors (e.g. same condition across several runs) is done by identifying the subject ID as the particle of the time-course filename before the first underscore character!** This means that VTC (or MTC) filenames **must be named** as follows: **Note: The combination of within-subjects regressors (e.g. same condition across several runs) is done by identifying the subject ID as the particle of the time-course filename before the first underscore character!** This means that VTC (or MTC) filenames **must be named** as follows:
Line 37: Line 37:
    
        options     optional 1x1 struct with fields        options     optional 1x1 struct with fields
-        .globsigs  add C components (of slicesas confounds (default: 0)+        .globsigs  add global signals as confound, one of 
 +                   0 - none 
 +                   1 - entire dataset (above threshold/within mask) 
 +                   2 - two (one per hemisphere, split at BV Z=128) 
 +                   3 or more, perform PCA of time courses and first N 
 +                   xff object(s), extract average time course from masks
         .ithresh   intensity threshold, default: 100         .ithresh   intensity threshold, default: 100
         .loadglm   boolean flag, load GLM file named in .outfile         .loadglm   boolean flag, load GLM file named in .outfile
         .mask      optional masking, default: no mask (for now only VTC)         .mask      optional masking, default: no mask (for now only VTC)
         .motpars   motion parameters (Sx1 cell array with sdm/txt files)         .motpars   motion parameters (Sx1 cell array with sdm/txt files)
 +        .motparsd  also add diff of motion parameters (default: false)
         .motparsq  also add squared motion parameters (default: false)         .motparsq  also add squared motion parameters (default: false)
         .ndcreg    if set > 0, perform deconvolution (only with PRTs!)         .ndcreg    if set > 0, perform deconvolution (only with PRTs!)
 +        .orthconf  orthogonalize confounds (and motion parameters, true)
         .outfile   output filename of GLM file, default: no saving         .outfile   output filename of GLM file, default: no saving
         .ppicond   list of regressors (or differences) to interact         .ppicond   list of regressors (or differences) to interact
         .ppirob    perform robust regression on VOI timecourse and remove         .ppirob    perform robust regression on VOI timecourse and remove
                    outliers from timecourse/model (threshold, default: 0)                    outliers from timecourse/model (threshold, default: 0)
-        .ppivoi    VOI used to extract time-course from+        .ppivoi    VOI object used to extract time-course from 
 +        .ppivoiidx intra-VOI-object index (default: 1)
         .prtpnorm  normalize parameters of PRT.Conds (true)         .prtpnorm  normalize parameters of PRT.Conds (true)
 +        .redo      selected subjects will be overwritten (default: false)
         .regdiff   flag, regress first discreet derivatives (diff) instead         .regdiff   flag, regress first discreet derivatives (diff) instead
         .restcond  remove rest condition (rest cond. name, default: '')         .restcond  remove rest condition (rest cond. name, default: '')
         .robust    perform robust instead of OLS regression         .robust    perform robust instead of OLS regression
 +        .savesdms  token, if not empty, save on-the-fly SDMs (e.g. '.sdm')
 +        .showsdms  token, passed to SDM::ShowDesign (if valid)
 +        .shuflab   PRT labels (conditions names) to shuffle
 +        .shuflabm  minimum number of onsets per label (1x1 or 1xL)
         .subsel    cell array with subject IDs to work on         .subsel    cell array with subject IDs to work on
         .tfilter   add filter regressors to SDMs (cut-off in secs)         .tfilter   add filter regressors to SDMs (cut-off in secs)
         .tfilttype temporal filter type (either 'dct' or {'fourier'})         .tfilttype temporal filter type (either 'dct' or {'fourier'})
         .vweight   combine runs/studies variance-weighted (default: false)         .vweight   combine runs/studies variance-weighted (default: false)
 +        .xconfound just as motpars, but without restriction on number
    
  Output fields:  Output fields:
Line 67: Line 81:
        regression models (to allow later continuation of crashed/broken        regression models (to allow later continuation of crashed/broken
        computation using the .subsel field)        computation using the .subsel field)
-       for .ppi to work, the model filenames must be PRTs!</file>+       for .ppi to work, the model filenames must be PRTs
 +       all additional fields for the call to PRT::CreateSDM are supported!</file>
  
 ===== Usage examples ===== ===== Usage examples =====
 Starting from scratch, the user has to do the following steps: Starting from scratch, the user has to do the following steps:
   * locate the filenames of the VTC (or MTC and SSM/TSM) files to be regressed   * locate the filenames of the VTC (or MTC and SSM/TSM) files to be regressed
-  * locate the filenames of the PRT (or SDM) files to be used as regressors (model)+  * locate the filenames of the [[xff - PRT format|PRT]] (or [[xff - SDM format|SDM]]) files to be used as regressors (model)
   * optionally locate the filenames of the realignment parameter files   * optionally locate the filenames of the realignment parameter files
   * make settings in the MDM structure   * make settings in the MDM structure
mdm.computeglm.txt · Last modified: 2011/04/03 01:21 by jochen