The first stage in data analysis is to preprocess the data.  This stage involves a series of manipulations done to the data which make later stages of analysis easier and provide improved results.  Each run for each subject must be preprocessed.  In order to provide the most options for analyzing our data, we actually split the preprocessing stream to produce two different preprocessed files for each run, one which is optimized for time series analysis and a second for using in GLM and deconvolution analyses.  The complete stream is illustrated below:

First, make "funcdata" folder under e.g. /data/VisualSearch/15VisualSearch/

cd into funcdata

run do_all_funcpreprocessing (see below)

	Shared Preprocessing Steps
	Step 1 - Reorienting the Data Step 2 - Slice Timing Correction Step 3 - Motion Correction Step 4 - Brain Extraction
Timeseries (Curv) Analysis Preprocessing Steps		GLM Analysis Preprocessing Steps
Step 5 - Temporal Filtering Step 6 - Mean Normalization Step 7 - Conversion to TES Format		Step 5 - Conversion to TES Format Step 6 - Spatial Smoothing
Output File Example		Output File Example
mfbrtStudy_Run1.tes		sbrtStudy_Run1.tes sbrtStudy_Run1_MoveParams.ref

Luckily for us, we have a single script, do_all_func_preprocessing, which will preprocess all of the individual runs for a subject with one command, insuring consistency throughout a study.  Click on the individual steps above to learn what each is doing and how, or find out how the do_all_func_preprocessing script works to automate it all.

After executing the do_all_func_preprocessing script, we will be left with two .tes files for each run (e.g. mfbrtStudy_Run1.tes and sbrtStudy_Run1.tes) as well as files containing the motion correction parameters.  The first of the .tes files is optimized for use in time series analysis.  The latter is optimized for use in GLM analyses and Deconvolution analyses.