Past, present, and future of adaptive rewiring in modelling neural self-organisation

Professor Cees van Leeuwen would like to introduce you to a research programme that himself and his lab members have been working on slowly but surely for the past 20 years. The project concerns the self-organisation of sparse neural networks through brain-inspired structural plasticity. Structural plasticity of the brain can be represented in highly simplified form as adaptive rewiring, the remodelling of connections according to the spontaneous dynamic synchrony in network activity.

Adaptive rewiring leads over time from initial random networks to brain-like complex network structures, i.e. modular small-world networks with rich-club effect. This basic principle was established 20 years ago, but in the following years, implementations have shifted from abstract models based on uniformly coupled oscillators to increasingly realistic models, without compromising the basic simplicity of the principle. The next 20 years could be focused on integrating self-organisation into models of evolution and development of the brain (a poor man's human brain project), or into artificial intelligence for exploring the functionality of adaptive rewiring in neural network sparsification.