bootstrap_SVT

purpose
usage
examples

Purpose:

This SVT command is used to statistically analyze the variations between groups of human brain functional volumes obtained under different stimulation paradigms.

You need to have all stereotactic data sets pre-processed according to the multi-subject part of the SVT manual. The output SVT volumes are 4 byte (floating point) per pixel stereotactic images of size 181x217y181z. The first image contains the positive and the second contains the negatively statistically significant voxels. Intensities of both SVT output images are in the range [0, infinity] and represent the exact Z-scores. This procedure uses a BOOTSTRAP approach to correct the intensity threshold level for the large number of statistical tests across the 3D volume.

Usage:

bootstrap_SVT 1 Num_points Pt_1_X_coord Pt_1_Y_coord Pt_1_Z_coord Sigma_1 ... Pt_N_X_coord Pt_N_Y_coord Pt_N_Z_coord Sigma_N Num_Files1gr A_1_Wp_2_T.img ... [[A_k_Wp_2_T.img]] R_1_Wp_2_T.img ... [[R_k_Wp_2_T.img]] SVT_vol.img

Usually we refer to the first paradigm as A (activation) and the second one as R (rest). Let A.img and R.img be the two volumes we are analyzing and let T.img be our anatomical reference (template, "avgedit.img"). We denote by A_2_R.img the warped-and-resliced (using any algorithm) A.img in R.img space. There are at least 3 different ways one can transform A.img and R.img in T.img space, see SVT manual.

In the above the following definitions are used:
1 just a flag! Num_points (N) the number of points we have manually identified, say on the SVT_simple volume, and we want to correct for multiple testing Pt_i_X_coord The X-coordinate (VOXEL coordinates, NOT world) for each 1 <= i <= N, selected point Pt_i_Y_coord The Y-coordinate for each 1 <= i <= N, selected point Pt_i_Z_coord The Z-coordinate for each 1 <= i <= N, selected point Sigma_i The estimate of the Regional standard deviation for each 1 <= i <= N. Obtained from the SVT_Simple-routine. These are the lobar stochastic estimates reported regionally by all SVT routines. Num_Files1gr The number of data files in each group (subtraction paradigm). A_1_Wp_2_T.img A(ctivation) paradigm data for subject 1, warped in T.img space A_k_Wp_2_T.img A(ctivation) paradigm data for subject k, warped in T.img space, 1<=k<=Num_Files1grp R_1_Wp_2_T.img R(est) paradigm data for subject 1, warped in T.img space R_k_Wp_2_T.img R(est) paradigm data for subject k, warped in T.img space, 1<=k<=Num_Files1grp SVT_vol.img One of the SVT (floating point) volumes produced, say, by SVT_simple

Examples:

SEE: bootstrap_SVT

bootstrap_SVT 1 2 14 36 78 2.4 56 75 120 4.56 3 grp1_data1 grp1_data2 grp1_data3 grp2_data1 grp2_data2 grp2_data3 SVT_pos.img

This command will use the post-processed groups of data (A's and R's are warped directly to the template, "avgedit.img") and the probabilistic partitioning information contained in "lobes9_diff_intens.img" to generate two multi-subject SVT output volumes (A_R_SVT_MS_Pos.img & A_R_SVT_MS_Neg.img) of the statistically significant differences between the two groups. Also look at the batch file "batch_SVT_MS". Note that the above example represents a single UNIX command-line.

)