Brain decoding of FMRI connectivity graphs using decision tree ensembles

  • Authors:

  • Jonas Richiardi;Hamdi Eryilma;Sophie Schwart;Patrik Vuilleumier;Dimitri Van De Ville

  • Affiliations:

  • Medical Image Processing Lab, Ecole Poly technique Fédérale de Lausanne, Lausanne and Medical Image Processing Lab, Université de Genève, Genève;Laboratory of Neurology and Imaging of Cognition, Université de Genèwe, Genève;Laboratory of Neurology and Imaging of Cognition, Université de Genèwe, Genève;Laboratory of Neurology and Imaging of Cognition, Université de Genèwe, Genève;Medical Image Processing Lab, Ecole Poly technique Fédérale de Lausanne, Lausanne and Medical Image Processing Lab, Université de Genève, Genève

  • Venue:
  • ISBI'10 Proceedings of the 2010 IEEE international conference on Biomedical imaging: from nano to Macro
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Functional connectivity analysis of fMRI data can reveal synchronized activity between anatomically distinct brain regions. Here, we exploit the characteristic connectivity graphs of task and resting epochs to perform classification between these conditions. Our approach is based on ensembles of decision trees, which combine powerful discriminative ability with interpretability of results. This makes it possible to extract discriminative graphs that represent a subset of the connections that distinguish best between the experimental conditions. Our experimental results also show that the method can be applied for group-level brain decoding.