Title: A Novel Metric Shows the Robustness of the Graph Communities to Brain-Tractography False-Positives

Abstract: We study the impact of the brain tractography false positives in the brain connectivity graphs. The representative input database for the analysis is the set of tractograms from the participants on the ISMRM-2015 Tractography Challenge. We propose 2 novel metrics to rank the quality of a tractogram when it is compared with known ground truth. The results of this study indicate that the estimation of graph communities is robust to high levels of overestimation in the connectivity.