Canonical cerebellar graph wavelets and their application to fMRI activation mapping

Summary, in English

Wavelet-based statistical parametric mapping (WSPM) is an extension of the classical approach in fMRI activation mapping that combines wavelet processing with voxel-wise statistical testing. We recently showed how WSPM, using graph wavelets tailored to the full gray-matter (GM) structure of each individual’s brain, can improve brain activity detection compared to using the classical wavelets that are only suited for the Euclidian grid. However, in order to perform analysis on a subject-invariant graph, canonical graph wavelets should be designed in normalized brain space. We here introduce an approach to define a fixed template graph of the cerebellum, an essential component of the brain, using the SUIT cerebellar template. We construct a corresponding set of canonical cerebellar graph wavelets, and adopt them in the analysis of both synthetic and real data. Compared to classical SPM, WSPM using cerebellar graph wavelets shows superior type-I error control, an empirical higher sensitivity on real data, as well as the potential to capture subtle patterns of cerebellar activity.

• Author: Hamid Behjat; Nora Leonardi; Leif Sörnmo; Dimitri Van De Ville
• Department/s: Department of Biomedical Engineering
• Publishing year: 2014
• Language: English
• Pages: 1039-1042
• Publication/Series: [Host publication title missing]
• Links: http://miplab.epfl.ch/pub/behjat1401.pdf; http://dx.doi.org/10.1109/EMBC.2014.6943771
• Document type: Conference paper
• Publisher: IEEE - Institute of Electrical and Electronics Engineers Inc.
• Topic: Medical Engineering
• Conference name: 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 2014
• Conference date: 2014-08-26 - 2014-08-30
• Conference place: Chicago, IL, United States
• Status: Published
• ISBN/ISSN/Other: ISSN: 1557-170X