Path ensembles and a tradeoff between communication efficiency and resilience in the human connectome

• Published in: Brain Structure and Function Vol. 222; no. 1; pp. 603 - 618
• Main Authors: Avena-Koenigsberger, Andrea, Mišić, Bratislav, Hawkins, Robert X. D., Griffa, Alessandra, Hagmann, Patric, Goñi, Joaquín, Sporns, Olaf
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.01.2017; Springer Nature B.V
• Subjects: Adult; Biomedical and Life Sciences; Biomedicine; Brain - anatomy & histology; Brain - physiology; Brain research; Cell Biology; Communication; Connectome - methods; Diffusion Magnetic Resonance Imaging; Female; Functional Laterality; Humans; Image Processing, Computer-Assisted; Male; Models, Neurological; Neural Pathways - anatomy & histology; Neural Pathways - physiology; Neurology; Neurosciences; Original Article; White Matter - anatomy & histology; White Matter - physiology; Young Adult; Communication resilience; Connectomics; Communication efficiency
• ISSN: 1863-2653; 1863-2661; 1863-2661; 0340-2061
• DOI: 10.1007/s00429-016-1238-5

Computational analysis of communication efficiency of brain networks often relies on graph-theoretic measures based on the shortest paths between network nodes. Here, we explore a communication scheme that relaxes the assumption that information travels exclusively through optimally short paths. The scheme assumes that communication between a pair of brain regions may take place through a path ensemble comprising the k-shortest paths between those regions. To explore this approach, we map path ensembles in a set of anatomical brain networks derived from diffusion imaging and tractography. We show that while considering optimally short paths excludes a significant fraction of network connections from participating in communication, considering k-shortest path ensembles allows all connections in the network to contribute. Path ensembles enable us to assess the resilience of communication pathways between brain regions, by measuring the number of alternative, disjoint paths within the ensemble, and to compare generalized measures of path length and betweenness centrality to those that result when considering only the single shortest path between node pairs. Furthermore, we find a significant correlation, indicative of a trade-off, between communication efficiency and resilience of communication pathways in structural brain networks. Finally, we use k-shortest path ensembles to demonstrate hemispherical lateralization of efficiency and resilience.