Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers

Poster Session A: Tuesday, August 12, 1:30 – 4:30 pm, de Brug & E‑Hall

Communication versus computation: The hidden costs, shaping the brain's architecture

Kayson Fakhar¹, Danyal Akarca², Andrea Luppi³, Stuart Oldham, Fatemeh Hadaeghi⁴, Petra Vertes, Claus Hilgetag, Duncan Astle; ¹University of Cambridge, ²Imperial College London, ³University of Oxford, ⁴University Medical Center Hamburg-Eppendorf

Presenter: Kayson Fakhar

Large-scale human brain networks exhibit complex topological characteristics, likely reflecting a balance among competing objectives such as minimizing wiring cost and maximizing communication efficiency. Interestingly, computational modelling has suggested that the connectivity of the brain is biased towards enhanced communication rather than a minimized wiring cost. Yet, the relationship between such communication efficiency and the functional capacity of the brain, e.g., to solve computational problems, remains unclear. To address this question, we used a game-theoretical framework in which individual brain regions establish connections only if it improves their signalling efficiency, given the wiring cost. We show that, firstly, complex network architectures naturally emerge from these local interactions, capturing some hallmarks of the brain. Secondly, resulting networks have both superior communication and reduced wiring cost compared to empirical brain networks. However, these optimal networks exhibited diminished memory capacity relative to empirical networks. Our findings suggest that efficient communication does not necessarily translate to improved computation. Instead, functional capacity may have played an essential role in shaping brain network architecture, potentially even at the expense of communication efficiency.

Topic Area: Brain Networks & Neural Dynamics

Extended Abstract: Full Text PDF