KULeuven automatisation scripts for SPM99


This section contains all scripts and functions for the automation of spm99.
In essence, only few adaptations have been made to the spm99b-scripts in order to implement the extra tools available in the final release of spm99.

The basic additions are:

1. Create an automation script for your experiment

To use the automated SPM99 functions you have to create a text file in which you supply all the information necessary to run the functions one after another. This text file will be executed in the matlab command window, and is thus itself a matlab script. This file should be stored in a directory in the MATLABPATH with the extention .m.
An example of a standard script for analysing fmri data from realignment to computation of statistics is provided with the functions and is called kul99example.m

Let us look at  such an automation script step by step.
The script file consists of several parts; comments start with the character %.

1.1 Obligatory part of the automation script

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Obligatory first part of the automation script                           %
%***************************************************************************
%  fmriTEST = 1 to run the simulation mode
%           = 0 to run the actual post-processing
%  fmriDIR  = the directory in which all your data is situated
%---------------------------------------------------------------------------
global fmriTEST fmriDIR;    %set global variabels
fmriTEST = 1;
fmriDIR  = '/images_rinda3/beatse/sidev6';
f99_CWD('LOG');             %make the LOG and RESULTS directory

This sets a few global variables to make the functions work.
The fmriTEST variable, if set to 1, makes the automation script run in a test or simulation mode: the script, if executed, will now simulate all steps of postprocessing and report warnings (e.g. missing image files). You will want to run the script in simulation mode to detect such warnings. If everything seems OK, you will change the fmriTEST value to 0. Then, if executed, the real thing happens...
The fmriDIR string is the name of the directory that holds your subdirectories with images. (on a windows PC it looks like 'c:\images_rinda\sunaert\spm99_kul_data')

Let's look how this part actually works: start a matlab, and start SPM99 as usual. Suppose you have copied the above script into a file called kul99example.m. Type in 'kul99example' into the matlab window. This window will now look like this:


 
 

1.2 Image processing steps in the analysis(i.e. slice timing correction, realignment, coregistration, normalisation and smoothing of the images).

1.3 Statistical part (i.e. constructing your design matrix and execution of the analysis).

1.4 Overlay of activations on anatomical slices
 



Webpage last updated on 26/7/2000. Authors: Stefan Sunaert & Erik Beatse.