Research > Software > MAST

Morphometric Atlasing Statistical Toolkit

The Morphometric Atlasing Statistical Toolkit is a collection of tools which deals with the general problem of creating statistical maps of data from populations of subjects.


  • Portable. Both R and Java are freely available on all commonly used operating systems.
  • Reads and writes many surface file formats. Mast uses the ShapeTools file format readers and writers, which support a growing number of popular surface file formats. These currently include: OpenDX General, OFF, VTK, Minc OBJ, PLY. Other file formats used in LONI are also supported, including: Ucf, CCBBM, Duff.
  • State of the art Statistics. Use of the R statistical language makes available a large number of high quality statistical computations.
  • Source code is available. All MAST source code is included. Source for libraries used (ShapeTools, R) is freely available for downloading.


The following applications are included:
A Java application that applies the False Discovery Rate theshold to a surface mesh file containing statistical p-values as vertex attributes.

A Java application that computes various measures of surface parameterization regularity.

A Java application that computes the Jacobian of an AIR Linear (affine) transformation.

The following libraries are included:
An R library providing R users to read and write common surface mesh file formats.It also implements commonly used statistical analyses.

A Java library which compares members of a group of shapes or mesh files.

A Java implementation of matrix logarithm and exponential functions.


System Requirements

  • OS: Any
  • Memory: 500 MB recommended for larger data set analyses.
  • Other: Java 1.5, or newer. rJava version 0.8-4.


The Java applications may be copied onto a local computer and used immediately. RShape requires that the R package installation procedures be run.


RShape allows the considerable power of R's statistical computations to be easily applied to neuroimaging data analysis. The ShapeComparison library allows rapid consistency checks to be applied to large number of surface mesh data files. These are typically performed prior to a computationally expensive data analysis.