Wavelet-based morphometry

WBM toolbox screenshot

What is wavelet-based morphometry?

Statistical analyses in voxel-based morphometry are usually based on either Gaussian random fields or in cluster-based statistics, with each of these approaches having its particular issues. Wavelet-based morphometry (WBM) is an alternative strategy consisting in conducting the statistical analysis (i.e., univariate tests) in the wavelet domain [1].

A well-known property of wavelets is that they are effective representing piecewise smooth images with few non-zero coefficients. Because of in the wavelet domain a small fraction of coefficients characterize the signal components, these coefficients can be identified and thus selected for the subsequent statistical analysis. This strategy allows reducing the number of hypothesis to be tested, thereby alleviating the multiple comparisons problem.

Download the WBM toolbox

What is the WBM toolbox for Matlab?

All the scripts and functions implemented in Ref. [1] were grouped and organized into a toolbox to facilitate the usage and evaluation of these tools by others research groups. The toolbox was implemented in Matlab ®, which is one of the more popular high-level technical computing languages. The toolbox is based on some functions implemented within SPM software package (i.e., SPM2, SPM5, or SPM8), as well as Wavelet-based SPM toolbox (http://bigwww.epfl.ch/publications/vandeville0704.html).

To install WBM in your computer:

  1. Check that Matlab and FSL are correctly installed, and that FSL can be executed from Matlab.
  2. Download the tarred gzip and extract the archive to a new directory.
  3. Add the main folder and all it's subdirectories to the Matlab path.
  4. At the Matlab prompt type: run_wbm
  5. If the GUI displayed at the upper right part of this page is visualized, then WBM has been correctly installed).

Example of output report

The toolbox creates a report in html with all the information of interest produced during the analysis. Please click here to see an example of the output report.


[1] Canales-Rodríguez EJ, Radua J, Pomarol-Clotet E, Sarró S, Alemán-Gómez Y, Iturria-Medina Y and Salvador R, "Statistical analysis of brain tissue images in the wavelet domain: Wavelet-based morphometry", Neuroimage 2013;72:214–226. DOI: 10.1016/j.neuroimage.2013.01.058