SGC_CWT

purpose
usage
examples

Purpose:

This WAIR command is used to obtain the Spread Group Classification (SGC) and the induced warp ranking having a family of warps applied to a group of data sets. For example, if we need to quantify warp performance, based on m subjects, each scanned under k different activation paradigms, and we wish to see large spread between the groups of m images, for the various functional activation stimuli, then we use the SGC analysis separately for every warp.

The SGC Classification of family of image-registration techniques is determined based on the warp performance on a group of data sets. You need to have executed "WT_IWT" and "wave_space_varThresh" first before you call "SGC_CWT", if wavelet-space analysis is to be done!

Usage:

SGC_CWT Num_Data TotNumParadigms TotNumBaseline Num_Dims x_1_Dim_Size [ [x_k_Dim_Size] ] [[WT_Data_1_Par_k]] [[WT_Data_2_Par_k]] ... [[WT_Data_m_Par_k]] results_WTData_SGC_whichWarp.text

where the following definitions are used:
Num_Data
the number of the data sets (m)
TotNumParadigms
the total number of activation paradigms, applied to all subjects (k)
TotNumBaseline
the total number of "base-line" activations. If no paradigm is being repeated, put 1
Num_Dims
the total number of dimensions of your data
x_1_Dim_Size
the the size of the first (fastest varying index) dimension of the data. Remember, all dimension sizes need to be powers of 2
x_k_Dim_Size
the the size of the k-th dimension of the data
WT_Data_i_Par_j
The (thresholded) wavelet transform of the data "i" under the activation paradigm "j", 1<=i<=Num_Data, 1<=j<=TotNumParadigms
results_WTData_SGC_whichWarp.text
file that will contain the output of the SGC analysis for the particular image registration

Examples:

./SGC_CWT 3 2 1 3 128 128 64 ./WTpet1_freq1_warp1_12812864.img ./WTpet1_freq2_warp1_12812864.img ./WTpet2_freq1_warp1_12812864.img ./WTpet2_freq2_warp1_12812864.img ./WTpet3_freq1_warp1_12812864.img ./WTpet3_freq2_warp1_12812864.img ./results_SGC_CWT_Warp1.text

This command will quantify, using SGC, the performance of 1 image registration (warping) based on 3 subjects each scanned twice, under two different activation paradigms. The data are 3D of matrix size 128x128y64z. The results of the SGC analysis will be saved in the output file "results_SGC_CWT_Warp1.text", in the current directory. Also look at the batch file "batch_SGC_CWT". Note that the above example represents a single command-line.

)