Changes in SDM version 6.23

The following information refers to changes in the behaviour or commands of SDM in version 6.23 as compared SDM version 5.15.

Input data files formats and SDM table file: The input data formats and sdm_table.txt file remain the same, except that sdm_table must contain a column t_thr indicating the threshold used to threshold each of the study maps. The column threshold is no longer required nor used.
Steps to perform while using the new method: The steps to perform now are preprocessing and mi/mi_lm (mean/linear model in the GUI) for a meta-analysis and perm for a FWE p-value correction. The FWE step may be conducted in several sessions.
Organization of study and analyses files: Intermediary and resulting files are now organized in folders. For an analysis names MyMean, data will be stored in folder analysis_MyMean. In the base folder of your study an xml file name sdmpsi_params.xml will be maintained with all the parameters for all the analysis performed in your study.
Preprocessing command: The preprocessing step and command remain practically the same. The only change is that now the parameter 'randomizations' is no longer required nor used, and an additional parameter, voxel size to use, is available. Additionally, a new mask for FreeSurfer studies has been added.
Mean: To compute a Mean there are now (up to) five parameters to select: Number of imputations to be used, number of concurrent threads to use, an optional filter and optionally one or two covariates (four if you're using the command-line)
Linear model (comparisons and meta-regressions): In SDM-PSI all linear models (comparisons and meta-regressions) are computed in one single command. Moreover, now the model (variables to use) and hypothesis need to be specified. Also, analogously to the computation of the Mean, for the Linear Model the number of imputations need to be specified. Note: Currently the intercept value is fixed to 0 for the computation of linear models.
Mask creation, extract and funnel plot: Now masks are created each in a determined analysis (and stored in analysis_NAME/masks/) and, therefore, an existing analysis must be selected while creating a mask. The same happens for extracts and funnel plots.