Both non-linear time-series and graph analysis have become powerful methods of investigating brain activity. In fact, both techniques have been successfully used in diagnosing the most important brain pathologies (e.g. epilepsy, Parkinson's, depression, Alzheimer's, etc.) or even in differentiating common mental tasks (e.g. logical operations, motor execution, states of sleep, etc.). Recently, several studies have shown that the dynamics of functional connectivity (FC) - commonly captured by the temporal evolution of similarities between brain regions' activities - defines an important marker of cognitive tasks or even pathological states, pointing to the need for merging these analytical paradigms. Bearing these aspects in mind, this research project aims to investigate the dynamics of the underlying functional connectivity from the perspective of Recurrence Quantification Analysis (RQA), an important paradigm of nonlinear dynamics theory. In particular, RQA stands out by characterizing the reappearance of states, providing access to invariant measures (correlation dimension, Kolmogorov-Sinai entropy), theoretical information measures (Rényi generalized entropies, mutual information, etc.) as well as to present sensitivity in detecting phase transitions and nonlinear relationships between coupled systems. From the practical standpoint, this project aims to use RQA in the characterization of FC dynamics for providing a new marker for movement intention underlying the motor imagery process, which may contribute for the design of more robust brain-computer interfaces. The project, which is part of the activities developed in the context of CEPID BRAINN (Brazilian Institute for Neuroscience and Neurotechnology) supported by FAPESP, also aims to establish a long-term cooperation between the proponent and the institution abroad, allowing the exchange of students, researchers and knowledge.
News published in Agência FAPESP Newsletter about the scholarship: