Measures for brain connectivity analysis: nodes centrality and their invariant patterns
- First Online:
- Cite this article as:
- Silva, L.M., Baltazar, C.A., Silva, C.A. et al. Eur. Phys. J. Spec. Top. (2017). doi:10.1140/epjst/e2016-60400-2
- 27 Downloads
The high dynamical complexity of the brain is related to its small-world topology, which enable both segregated and integrated information processing capabilities. Several measures of connectivity estimation have already been employed to characterize functional brain networks from multivariate electrophysiological data. However, understanding the properties of each measure that lead to a better description of the real topology and capture the complex phenomena present in the brain remains challenging. In this work we compared four nonlinear connectivity measures and show that each method characterizes distinct features of brain interactions. The results suggest an invariance of global network parameters from different behavioral states and that more complete description may be reached considering local features, independently of the connectivity measure employed. Our findings also point to future perspectives in connectivity studies that combine distinct and complementary dependence measures in assembling higher dimensions manifolds.