lunduniversity.lu.se

Biomedical Engineering

Faculty of Engineering, LTH

Denna sida på svenska This page in English

Software

Toolbox for constructing signal-adapted systems of spectral kernels for analysing graph signals

SPG (signal processing on graphs) package - v1.00 (July 2016)

This package contains code for constructing signal-adapted systems of spectral kernels for a given graph and a given graph signal set, based on theory presented in:

[1] Hamid Behjat, Ulrike Richter, Dimitri Van De Ville, Leif Sornmo, ''Signal-adapted tight frames on graphs'', IEEE Trans. Signal Process., 2016, doi: 10.1109/TSP.2016.2591513.

The toolbox is implemented in MATLAB. 

The associated paper can be downloaded from either here or here

The toolbox can be downloaded from either here or here.

Other features of the toolbox include:

  • The Minnesota road graph, the Alameda graph and the cerebellum graph, together with their associated datasets as described in the paper are also included in the package. The cerebellar dataset is partially included due to its large size, but the remainder can also be provided upon request. 
  • Demos for constructing signal-adapted systems of spectral kernels using your own graph and graph signal set, and to decompose a set of graph signals using the constructed frame (not necessarily the same signals used to construct the frame) are also provided.
  • The functions and demos also provide means to construct SGWT frame [2], Meyer-like [3] and spectrum-adapted [4] systems of spectral kernels for comparison with the proposed signal-adapted systems of spectral kernels.

[2] Hammond, et al., ''Wavelets on graphs via spectral graph theory'', Appl. Comput. Harmon. Anal., vol. 30, pp. 129-150, 2011.

[3] Leonardi, et al., ''Tight wavelet frames on multislice graphs'', IEEE Trans. Signal Process., vol. 61(13), pp. 3357-3367, 2013.  

[4] D. I. Shuman, et al., ''Spectrum-adapted tight graph wavelet and vertex-frequency frames'', IEEE Trans. Signal Process., vol. 63(16), pp. 4223-4235, 2015.

Toolbox for enhanced fMRI activation mapping using anatomically-adapted graph wavelets

gwSPM : graph-based, wavelet-based Statistical Parametric Mapping Toolbox 

This toolbox contains code implementing the framework presented in the following paper: 

H. Behjat, N. Leonardi, L. Sörnmo, D. Van De Ville, Anatomically-adapted graph wavelets for improved group-level fMRI activation mapping,” NeuroImage, vol. 123, pp. 185-199, 2015,

which can be downloaded from either here or here

The toolbox is implemented in MATLAB, and is developed as a toolbox for the SPM software package. 

For an overview of the toolbox, the poster which was presented at the annual meeting of the Organization for Human Brain Mapping (OHBM, June 2018, Singapore) can be viewed here

Few snapshots of the toolbox can be seen in the below images.

 

The toolbox can be downloaded from either here or here