#!/bin/tcsh -f # ver=3.0 # Calculate ReHo of the LFF time series-- the Kendall Coefficient of # Concordance (KCC, or Kendall's W) of a neighborhood of time series. # Three examples are given here, based on different possibilities of # 'neighborhood' definition: # A): for each voxel, take face-, edge- and corner-wise neighbors, for # a total of 27 voxels # B): ellipsoidal shape of 'hood, defined by desired number of voxels # each semi-axis (this could be useful for non-isotropic voxels, for # instance, even though the sample set voxels here are isotropic) # C): calculate ReHo of networks of arbitrarily shaped ROIs via labelled # masks-- the input time series is the un-LFF-bandpassed time series # that had been mapped to DWI space (same space as the 3dROIMaker'ed # networks) # also, chi_sq values are chosen to be output here as well. # Case A, default 3dReHo \ -inset FMRI/REST_filt_LFF+orig \ -prefix FMRI/REST_REHO_N27 \ -chi_sq \ -overwrite # Case B 3dReHo \ -inset FMRI/REST_filt_LFF+orig \ -neigh_X 3 \ -neigh_Y 4 \ -neigh_Z 2 \ -prefix FMRI/REST_REHO_ELL \ -chi_sq \ -overwrite # Case C-- NB: a map of `case A' values is also calculated by default. # The ROI network values are output in *.vals and *.chi text files, # one network per line (in order of subbrick). 3dReHo \ -inset REST_in_DWI.nii.gz \ -mask mask_DWI+orig \ -in_rois ROI_ICMAP_GM+orig \ -prefix FMRI/REST_in_DWI_REHO_ROI \ -overwrite