Hi ptaylor,
Thanks once again for the help. We're getting much closer to resolved!
Re 1): I ran those pieces of code you provided and here are the following outputs:
Running the following code on a sample subject data file in original space...
3dinfo -ad3 DATASET
...provides:
Quote
2.750000 2.750000 5.000000
3dinfo -n4 DATASET
...provides:
Quote
80 80 26 1
3dinfo DATASET
...provides:
Quote
++ 3dinfo: AFNI version=AFNI_19.2.24 (Sep 13 2019) [64-bit]
Dataset File: DW+orig
Identifier Code: AFN_Wz_yXl6e2Jv-HGG0NxFH6Q Creation Date: Fri Jun 24 15:34:38 2022
Template Space: ORIG
Dataset Type: Intensity (-fim)
Byte Order: LSB_FIRST [this CPU native = LSB_FIRST]
Storage Mode: BRIK
Storage Space: 665,600 (666 thousand) bytes
Geometry String: "MATRIX(2.75,0,0,-111.1003,0,2.75,0.04542,-112.6167,0,-0.024981,5,-12.63179):80,80,26"
Data Axes Tilt: Oblique (0.521 deg. from plumb)
Data Axes Approximate Orientation:
first (x) = Right-to-Left
second (y) = Anterior-to-Posterior
third (z) = Inferior-to-Superior [-orient RAI]
R-to-L extent: -111.100 [R] -to- 106.150 [L] -step- 2.750 mm [ 80 voxels]
A-to-P extent: -112.617 [A] -to- 104.633 [P] -step- 2.750 mm [ 80 voxels]
I-to-S extent: -12.632 -to- 112.368 [S] -step- 5.000 mm [ 26 voxels]
Number of values stored at each pixel = 1
-- At sub-brick #0 '#0' datum type is float: -3.12815 to 3.19592
So all together it looks like my original voxel size is 2.75 X 2.75 X 5 mm3. Which means my voxel size was similar to the example data set you mentioned earlier at 2.5mm, yet the example data set you mentioned had 68,809 voxels whereas my 70% intersection mask contains ~1.7 million voxels. Do you think something possibly went wrong in the processing pipeline that would lead to the ~1.7 million voxels I'm seeing in my 70% intersection mask?
Re 4): My predecessors did not use afni_proc.py to do the pre-processing. Rather, they used an old guide that was provided by researchers from Baycrest. So, I do not have the QC output from afni_proc.py. Is there a way I can obtain that without needing to redo everything?
If it helps, I did take your suggestion and checked that the intersection mask seems okay visually. We are actually not using an anatomical template, but rather an average anatomical image from the 18 subjects. I took some pictures to show you (see attached images below) that show the 70% intersection mask seems to fit well within the average anatomical image and when scrolling through the 3 planes in AFNI GUI I did not notice anything obvious missing or extending out past the skull.
Re 5): That makes complete sense now, thanks! Since we are comparing two conditions, I will stick with single, bisided tests without any conversion between tables.
