Meeting Banner
Abstract #3626

A Random-Walk Driven Segmentation of Resting State fMRI Data: Evaluation of Visual Cortex Sub-Communities is Enhanced by Physiological Noise Correction

Tommaso Gili1, Ibrahim Eid2, Kevin Murphy1, Ashley Harris1, Guido Caldarelli3, Bruno Maraviglia2, Richard Geoffrey Wise1

1Cardiff University Brain Research Imaging Centre (CUBRIC), School of Psychology, Cardiff University, Cardiff, Wales, United Kingdom; 2Dipartimento di Fisica, Universit di Roma Sapienza, Roma, Italy; 3CNR-ISC Dipartimento di Fisica, Univerist di Roma Sapienza, Roma, Italy

Parcellation of the cortex into individual subunits based on correlations in resting-state data opens up the possibility of developing a subunit atlas analogous to the Brodmann areas but based on cortical function rather than cytoarchitecture. Graph theory is a common methodology for studying complex networks. Graph-based clustering approaches have also been applied to the analysis of brain networks using resting-state fMRI. Here we used a random-walk-based algorithm to parcellate visual cortex in healthy subjects in a resting-state condition. Moreover we demonstrate the need for physiological noise removal to obtain consistent results across subjects from this functional segmentation method.