afni_ci_test_data/AFNI_data6/roi_demo/roi_guide.txt

HANDS-ON EXAMPLES FOR ROI DEMO
---------------------------------
---------------------------------

	cd AFNI_data6/roi_demo

-------------------------------------------------------------------
-------------------------------------------------------------------

3dresample:
RESAMPLE ROI MASK FROM HIGH RES TO LOW RESOLUTION

(1)	3dresample -master rall_vr+orig -prefix anat_roi_resam \
		   -inset anat_roi+orig -rmode NN

-------------------------------------------------------------------
-------------------------------------------------------------------

3dmaskave:
USE THE RESAMPLED MASK ON A 3D+TIME DATASET  TO COMPUTE A MEAN VALUE FOR THE
VOXELS THAT FALL WITHIN THE MASK AT EACH TIMEPOINT (124 TIMEPTS IN THIS E.G.)

(2a)	3dmaskave -mask anat_roi_resam+orig -q rall_vr+orig > epi_avg.1D

	
(2b)	1dplot -yaxis 1000:1200:2:1 epi_avg.1D

-------------------------------------------------------------------
-------------------------------------------------------------------

3dmaskdump:
USE THE RESAMPLED MASK ON A FUNCTIONAL DATASET.  IN THIS E.G., WE ARE SELECTING
A SPECIFIC SUB-BRICK IN OUR FUNC DSET (e.g., SUB-BRICK #2, T-VALUES FOR "VISUALLY 
RELIABLE" (Vrel) STIMULi) AND 'DUMPING' ONLY THE T-STATS FOR 'Vrel' FOR THE VOXELS 
THAT FALL WITHIN THE ROI MASK

(3a)	3dmaskdump -noijk -mask anat_roi_resam+orig 'func_slim+orig[2]' \
		   > Vrel-tstats.txt


(3b)	more Vrel-tstats.txt


NOTE: YOU CAN DO THIS WITH MORE THAN 1 SUB-BRICK AT A TIME. E.G.,

      
(3c)    3dmaskdump -noijk -mask anat_roi_resam+orig	         \
		   'func_slim+orig[1,2]' > Vrel_beta_t.txt

(3d)	more Vrel_beta_t.txt

-------------------------------------------------------------------
-------------------------------------------------------------------

3dROIstats:
COMPUTE A MEAN FOR VOXELS IN AN ROI MASK.  CAN BE ANAT ROI OR FUNC ROI.
DIFFERENT FROM 3DMASKAVE
	  - WITH 3dmaskave, IF YOU HAVE A MASK WITH >1 ROI IN IT, THIS PROGRAM
	    WILL COLLAPSE THOSE ROIS AND COMPUTE ONE GRAND MEAN.
	  - WITH 3dROIstats,IF YOU HAVE A MASKE WITH >1 ROI IN IT, THIS PROGRAM
	    WILL COMPUTE INVIDIDUAL MEANS FOR EACH ROI.

	 
(4)	3dROIstats -mask 3rois+orig  func_slim+orig.'[0]'

-------------------------------------------------------------------
-------------------------------------------------------------------

CREATE A FUNCTIONAL ROI MASK, USING A FUNC DATASET AS OUR GUIDE (RATHER THAN
AN ANATOMICAL UNDERLAY)

YOU MUST OPEN THE FUNC DSET, SELECT THE SUB-BRICK OF INTEREST, SET A THRESHOLD 
(TO KEEP ONLY THE CLUSTERS THAT SURVIVE THE THRESHOLD) & DECIDE HOW BIG YOU
WANT THOSE FUNC CLUSTERS TO BE.

3dclust:     
3DCLUST SHOWS YOU A LIST OF CLUSTERS THAT SURVIVE THE CRITERIA YOU'VE SET
	      -1clip sets threshold the user selects, e.g.,   F>=99, 
	      rmm=1 (voxels have to touch to be part of the cluster), 
	      vmul 200 (1x1x1x200=200))

(5a)	3dclust -dxyz=1 -1clip 99  1.0  200 func_slim+orig.'[0]'



3dmerge:
3DMERGE CREATES AN AFNI DATASET OF THE CLUSTERS THAT HAVE SURVIVED THE
CRITERIA YOU'VE SET

(5b)	 3dmerge -prefix func_roi  -dxyz=1 -1clip 99 -1clust_order 1 200 \
		 func_slim+orig.'[0]'

-------------------------------------------------------------------
-------------------------------------------------------------------

whereami:
CREATE AN ROI DATASET USING VARIOUS ATLASES AVAILABLE WITH THE AFNI PROGRAM whereami
(BUT FIRST CREATE A CLUSTER WITH 3DCLUST AND SAVE IT AS A 1D (TXT) FILE):

(6)	3dclust -dxyz=1 -1clip 99 1 200 func_fullF+tlrc > clusts.1D
	
	whereami -coord_file clusts.1D'[1,2,3]' -tab | less

-------------------------------------------------------------------
-------------------------------------------------------------------

whereami:
EXTRACT ROIS FOR VARIOUS ATLAS REGIONS USING SYMBOLIC NOTATION

(7)	whereami -mask_atlas_region TT_Daemon:left:amy
		
-------------------------------------------------------------------
-------------------------------------------------------------------

whereami:
REPORT ON THE OVERLAP OF ROIS WITH ATLAS-DEFINED REGIONS

(8)	whereami -omask anat_roi+tlrc

-------------------------------------------------------------------
-------------------------------------------------------------------

more on ROIs:
SPECIFY ATLAS-BASED ROI MASKS DIRECTLY LIKE THIS:

(9a)	3dmaskave -mask TT_Daemon:left:hippocampus func_fullF_1mm+tlrc

(9b)	3dcalc -a 'CA_N27_ML::hippo' -b 'func_fullF_1mm+tlrc' \
	       -expr '(step(a)*b)' -prefix nice_roi

(9c)	3dcalc -a 'TT_Daemon::amygdala' -b 'CA_N27_ML::amygdala' \
	       -expr 'step(a)+2*step(b)' -prefix compare.maps