Axis 1 : Neuronal dynamics
A central aim of our team is to understand the neuronal dynamics that underlie the resting activity of the brain in normal and pathological conditions. A particular model of interest is tinnitus and its associated neuronal dynamics, since the phantom auditory perceptions of tinnitus are not linked to sound stimulations, but to abnormal, spontaneous neuronal activity. This model offers important perspectives in the application of more theoretically-oriented studies of neuronal dynamics. These studies extend our previous work on the neural correlates of tinnitus, which focused on the characterization of more static properties of neural firing, namely the firing rate, synchrony (spiking activity cross-correlation, coherence), and bursting activity.
The central approach of our theoretical studies of neuronal dynamics is based on the fact that if biophysics of the different recording techniques are modeled in a detailed manner, the quality of the information obtained at each scale of observation of the nervous system are still ill-defined. We wish to extend our multi-scale approach of electroencephalography (EEG) in a larger framework to obtain valid criteria in measuring neuronal dynamics related to different scales. This approach is founded on the methods of discretization of dynamical systems obtained by approximations, partitions or « coarse-graining », studied in nonlinear physics. The general principle will be to consider brain dynamics as a dynamical systems of very high dimension, which can be defined either in terms of local degrees of freedom (such as single cell recordings) or as measurements over several scales (such as local field potentials or EEG). The problem to solve is to define what information loss is inherent to scale changes and under which conditions these losses can be minimized.
The multi-scale methods developed in theoretical studies will be applied to the study of emotion regulation using EEG or near-infrared spectroscopy (NIRS) in collaboration with K. Doba and J.-L. Nandrino (SCALab, UMR 9193 Université de Lille, CNRS). We will extend our previous studies of behavioral emotional interactions by focusing on both the empirical characterization of dyadic interactions between mother and baby, and the modeling and dynamics of emotional expressions in autobiographical stories of healthy subjects and patients (anorexia, schizophrenia).
Finally, we will also evaluate the possibility of transferring the various techniques developed under this collaboration to clinical situations of patients with disabling hearing disorders, such as tinnitus and hyperacusis.