Impact of normalization and segmentation on brain graphs

22 Graph theory has recently received a lot of attention from the neuroscience community as a method to 23 represent and characterize brain networks. Still, there is a lack of a gold standard for the methods that 24 should be employed for the preprocessing of the data and the construction of the networks, as well as 25 a lack of knowledge on how different methodologies can affect the metrics reported. We used graph 26 theory analysis applied to resting-state functional Magnetic Resonance Imaging (rs-fMRI) to 27 investigate the influence of different node-defining strategies and the effect of normalizing the 28 functional acquisition on several commonly reported metrics used to characterize brain networks. The 29 nodes of the network were defined using either the individual FreeSurfer segmentation of each subject 30 or the FreeSurfer segmented MNI (Montreal National Institute) 152 template, using the Destrieux and 31 sub-cortical atlas. The functional acquisition was either kept on the functional native space or 32 normalized into MNI standard space. The comparisons were done at three levels: on the connections, 33 on the edge properties and on the network properties levels. Our results reveal that different 34 registration and brain parcellation strategies have a strong impact on all the levels of analysis, 35 possibly favoring the use of individual segmentation strategies and conservative registration 36 approaches. In conclusion, several technical aspects must be considered so that graph theoretical 37 analysis of connectivity MRI data can provide a framework to understand brain pathologies.