A Predictive Coding Perspective of Brain Dynamics: the case of Oscillatory Travelling Waves

One of the most exciting yet puzzling questions in Neuroscience is how the brain coordinates the activity between different areas, integrating distinct representations into conscious percepts and thoughts. For decades neuroscientists have investigated how the brain orchestrates diverse regions activity, pointing at oscillations as one of the key mechanisms involved in such a process. However, previous research has mainly focused on the temporal aspect of oscillatory dynamics, largely overlooking how oscillations propagate through the brain. Although rhythmic traveling waves have recently gained renewed interest, their functional role and relation to cognitive functions remain largely unknown. In this project, I will address this fundamental question: what is the role of oscillatory traveling waves in brain dynamics? I plan to take on this challenge using a multi-scale computational approach, modeling neural dynamics within and between cortical regions, as well as cortical-thalamic interactions. Importantly, the novelty of this approach consists in framing the model in the light of Predictive Coding principles, to test the compelling yet striking hypothesis that traveling waves encode Predictions and Prediction-Errors. The results of the simulations will be compared against experimental recordings in human participants to validate and assess the models predictions. Lastly, some implementations will turn into deep learning architectures, to test their dynamics in visual tasks while improving current models of artificial vision. All in all, this proposal can significantly advance our understanding of the neurophysiological mechanisms involved in sensory and cognitive functions, testing whether and how oscillatory traveling waves are a critical mechanism in neural dynamics, and producing fundamental results in the scientific field and future technological applications.