Computer Information for Image Processing and Analysis
All imaging procedures were completed on the Fulton Supercomputer hosted at Brigham Young University. The Fulton Supercomputing Lab offers 21,552 CPU cores, 78TB of memory, and 972 compute nodes. All nodes run Red Hat Enterprise Linux 6.6. For further details, please visit https://marylou.byu.edu/documentation/resources.
MRI Acquisition
Image acquisition. Neuroimaging data were acquired on a 3-Tesla Siemens Trio Tim whole-body MR scanner (Siemens) using a 32-channel head coil for reception. Each scan session began with a three-plane localizer, and total acquisition time was approximately 30 minutes. A 7-minute T1-weighted MPRAGE structural scan, included contiguous 176 sagittal slices with the following additional parameters: TR = 1900 ms, TE = 2.26 ms, FOV = 250 mm, acquisition matrix = 256 x 215 mm, flip angle = 9o, voxel size = 1.0 mm3. and slice thickness = 1.0 mm
MRI Preprocessing
The first step in processing our data was to convert the DICOMs to NIfTI files using dcm2niix v1.0.20170818 downloaded from
https://github.com/rordenlab/dcm2niix.git. Running dcm2niix requires Parallel Implementation of GZip (pigz) version 2.3.4 from
https://github.com/madler/pigz.git. Anonymization of identifying information is included as part of this step. The code for this step is: