Conférence de Romain Goutagny "Decoding the hippocampal (and cortical) oscillatory Complexity to predict behavior"

Résumé de la conférence

"Theta and gamma oscillations are believed to organise hippocampal activity. The current dominant view holds that two gamma frequency sub-bands occur at different theta phases in CA1, are produced by different generators and sub-serve different cognitive operations. Recent evidence suggested however that most individual theta cycles actually contain very diverse and complex gamma contents. We thus aimed at exploring further this theta-gamma coupling diversity by using a new topological approach based on the detection of several individual gamma bouts per theta cycle. We performed an exhaustive analysis, theta cycle by theta cycle, of the different theta-gamma motifs found in all hippocampal CA1 sublayers in mice during the completion of spatial memory task. Strikingly, we found that gamma bursts can be found for nearly every combination of frequency, amplitude and theta phase in all hippocampal CA1 layers, with less than 5% of the cycles actually matching the classical template motif. Importantly, the underlying complexity of theta-gamma coupling considered at the theta cycle timescale seems nonetheless to carry behavioural information: running speed and coarse position within the maze could be predicted by a machine learning algorithm from these seemingly stochastic properties of isolated theta cycles. Altogether, our results clearly point toward the existence of an interdependence between behavior, learning and the rich, complex and largely overlooked diversity of hippocampal theta-gamma oscillations."