Étude des systèmes complexes : des réseaux au connectome du cerveau
|Advisor:||Dubé, Louis J.; Côté, Daniel; Desrosiers, Patrick|
|Abstract:||Connectomics is the study of the brain connectivity maps (animal or human), described as complex networks and named connectomes. The organization of the connections, including the network’s hidden hierarchy, plays a major role in our understanding of the functional and structural complexity of the brain. Until now, the hierarchical models in connectomics have exhibited few emergent properties and have proposed regular structures whereas conectomes and real networks show complex structures. We introduce a new growth model of hierarchical networks based on preferential attachment (HPA - hierarchical preferential attachment). The structure can be controlled by a small set of parameters to fit real networks. We show how functional properties emerge from the projection of the hierarchical organization. Furthermore, we use HPA to investigate the minimum level of activity of the brain. The network response under binary dynamics shows evidence of persistent activity, similar to the resting-state of the brain. Even though hierarchical organization is beneficial for sustained activity, we show that persistent activity emerges from the control of the structure over the dynamics.|
|Document Type:||Mémoire de maîtrise|
|Open Access Date:||24 April 2018|
|Collection:||Thèses et mémoires|
All documents in CorpusUL are protected by Copyright Act of Canada.