CONN extensions:

PREP

CONN's PREP module provides streamlined access to all of CONN's preprocessing (e.g. realignment, normalization, etc.) and denoising steps (e.g. scrubbing, aCompCor) to prepare your functional/anatomical data for arbitrary (non-connectivity) analyses

The PREP module can be run by first defining the source of data and details of your preprocessing pipeline in a .cfg file like the following (for CONN's default minimal preprocessing pipeline):

config_preproc01.cfg

#functionals

/data/func/func_run01.nii

/data/func/func_run02.nii


#structurals

/data/anat/anat01.nii


#steps

default_mni


#RT

2


#sliceorder

interleaved (Siemens)


example preprocessing .cfg file (see preprocessing .cfg documentation for additional details of this file format)

and then simply using the following Matlab syntax to run that preprocessing/denoising pipeline:

>> conn_module PREP /data/config_preproc01.cfg

(note: if using CONN standalone version, use the OS command "conn module ###" in place of any Matlab command of the form "conn_module ###")

The results of a preprocessing/denoising pipeline are stored in the same directory as your original functional/anatomical data following standard SPM prefix conventions for file names, for example:

example output files after a standard preprocessing pipeline (original functional data rest.nii, fully preprocessed data swaurest.nii)

In addition to preprocessing functional and anatomical data, PREP can also be used to denoise functional data. Denoising steps in PREP include linear regression of confounding effects (e.g. scrubbing, and anatomical CompCor) as well as band-pass filtering. These steps are run using the same procedures as above and simply including the functional_regression and/or functional_bandpass steps as part of your configuration file. For example, the configuration file below defines a standard preprocessing and denoising pipeline that includes CONN's minimal preprocessing pipeline, plus anatomical CompCor (5 White Matter and 5 Gray Matter components), scrubbing (using default outlier detection thresholds), regression of motion effects and their first-order derivatives, and band-pass filtering (0.01 to 0.10Hz):

config_preproc02.cfg

#functionals

/data/func/func_run01.nii

/data/func/func_run02.nii


#structurals

/data/anat/anat01.nii


#steps

default_mni

functional_regression

functional_bandpass


#RT

2


#sliceorder

interleaved (Siemens)


#reg_names

realignment

scrubbing

White Matter

CSF


#reg_dimensions

inf

inf

5

5


#reg_deriv

1

0

0

0


#bp_filter

0.01 0.10

example preprocessing .cfg file (see preprocessing .cfg documentation for additional details of this file format)