Includes reading in and viewing DICOM images, converting to NIfTI, adjusting voxel intensity values, performing and evaluating different inhomogeneity correction methods, skull stripping/brain extraction, automatic tissue segmentation, atlas registration, Gaussian smoothing, and white and gray matter volume estimation.
The only download required before starting is FSL software. All other downloads—including retrieving sample data—occur within the R script.
This script is meant for a single case, but it can be modified easily to construct a batch processing pipeline.
oro.dicom, oro.nifti, devtools, neurobase*, cmaker*, ITKR*, ANTsR*, extransr*, papayar*, fslr*, reshape2, ggplot2, qpdf, viridis
*Hosted on GitHub.
FSL software is required.
BRAINIX sample data is available through OsiriX.
Reference atlas (MNI 152 T1w, 0.5mm, asymmetrical) can be retrieved from NIST.
Tested on a Mac running OS Monterey 12.1 using R version 4.0.2. Several utilized R packages hosted on GitHub are configured for Mac or Linux. If running R version 4.1 or later, it is highly recomended to switch active R version to 4.0 using, e.g., RSwitch.