BrainSuite is a magnetic resonance (MR) image analysis tool designed for identifying tissue types and surfaces in MR images of the human brain. The major functionality of these tools is to extract and parameterize the inner and outer surfaces of the cerebral cortex and to segment and label gray and white matter structures. BrainSuite also provides several tools for visualizing and interacting with the data.
BrainSuite is collaboratively developed between David Shattuck at the Laboratory of Neuro Imaging (LONI), UCLA and Richard Leahy in the Biomedical Imaging Group, USC. Other contributors to the development of the methods embodied in this software include: Anand A. Joshi, Belma Dogdas, Stephanie Sandor-Leahy, Bijan Timsari, and Paul M. Thompson.
The currently released version of the software (BrainSuite11a) can be dowloaded here. This release includes both the graphical user interface version, available for Windows platforms, and command line versions of the the routines used for cortical surface extraction (available for Mac OSX, and Linux). The GUI can also be run on Linux and Mac OSX using WINE, Darwine (Mac OSX), or Crossover (commercial), or using virtual machine software such as Virtual Box or VMWare Fusion
Current BrainSuite features include:
- Visualization tools, such as MRI visualization in 3 orthogonal views (either seperately or in 3D view), and overlayed surface visualization of cortex, skull, and scalp
- Cortical surface extraction, using a multi-stage user friendly approach. Tools include brain surface extraction, bias field correction, voxel classification, cerebellum removal, and surface generation
- Topological correction of cortical surfaces, which uses a graph-based approach to remove topological defects (handles and holes) and ensure a tessellation with spherical topology
- Interactive labeling of sulcal landmarks in surface and orthogonal views
- Skull and scalp surface extraction
A new version of BrainSuite will be released soon that will include coregistration of cortical surfaces to an atlas and surface-constrained volume registration to an atlas that produces a one-to-one mapping between cortical surfaces.