Data-driven discovery of canonical large-scale brain dynamics

The Hopf whole-brain model and its linear approximation

Whole-brain models have proven to be useful to understand the emergence of collective activity among neural populations or brain regions. These models combine connectivity matrices, or connectomes, with local node dynamics, noise, and, eventually, transmission delays. Multiple choices for the local dynamics have been proposed. Among them, nonlinear oscillators corresponding to a supercritical Hopf bifurcation have been used to link brain connectivity and collective phase and amplitude dynamics in different brain states. Here, we studied the linear fluctuations of this model to estimate its stationary statistics, i.e., the instantaneous and lagged covariances and the power spectral densities. This linear approximation-that holds in the case of heterogeneous parameters and time-delays-allows analytical estimation of the statistics and it can be used for fast parameter explorations to study changes in brain state, changes in brain activity due to alterations in structural connectivity, and modulations of parameter due to non-equilibrium dynamics.

A low dimensional embedding of brain dynamics enhances diagnostic accuracy and behavioral prediction in stroke

Large-scale brain networks reveal structural connections as well as functional synchronization between distinct regions of the brain. The latter, referred to as functional connectivity (FC), can be derived from neuroimaging techniques such as functional magnetic resonance imaging (fMRI). FC studies have shown that brain networks are severely disrupted by stroke. However, since FC data are usually large and high-dimensional, extracting clinically useful information from this vast amount of data is still a great challenge, and our understanding of the functional consequences of stroke remains limited. Here, we propose a dimensionality reduction approach to simplify the analysis of this complex neural data. By using autoencoders, we find a low-dimensional representation encoding the fMRI data which preserves the typical FC anomalies known to be present in stroke patients. By employing the latent representations emerging from the autoencoders, we enhanced patients' diagnostics and severity classification. Furthermore, we showed how low-dimensional representation increased the accuracy of recovery prediction.

Characterization of regional differences in resting-state fMRI with a data-driven network model of brain dynamics

Model-based data analysis of whole-brain dynamics links the observed data to model parameters in a network of neural masses. Recently, studies focused on the role of regional variance of model parameters. Such analyses however necessarily depend on the properties of preselected neural mass model. We introduce a method to infer from the functional data both the neural mass model representing the regional dynamics and the region- and subject-specific parameters while respecting the known network structure. We apply the method to human resting-state fMRI. We find that the underlying dynamics can be described as noisy fluctuations around a single fixed point. The method reliably discovers three regional parameters with clear and distinct role in the dynamics, one of which is strongly correlated with the first principal component of the gene expression spatial map. The present approach opens a novel way to the analysis of resting-state fMRI with possible applications for understanding the brain dynamics during aging or neurodegeneration.