Technical Program

Paper Detail

Paper: PS-1B.31
Session: Poster Session 1B
Location: Symphony/Overture
Session Time: Thursday, September 6, 18:45 - 20:45
Presentation Time:Thursday, September 6, 18:45 - 20:45
Presentation: Poster
Paper Title: A dynamical systems model of intrinsic and evoked activity, variability, and functional connectivity
Manuscript:  Click here to view manuscript
Authors: Takuya Ito, Brian Keane, Ravi Mill, Richard Chen, Luke Hearne, Katelyn Arnemann, Rutgers University, United States; Biyu He, New York University, United States; Horacio Rotstein, New Jersey Institute of Technology, United States; Michael Cole, Rutgers University, United States
Abstract: Neural signals can be measured experimentally by estimating levels of brain activity, variability, and functional connectivity. However, these neural measures have often been studied independently from one another, making it difficult to infer precise underlying causes of the phenomena. Here we provide a mechanistic framework that relates activity, variability, and functional connectivity in neural mass models. We hypothesized that statistical estimates of activity, variability, and functional connectivity are emergent properties describing network interactions governed by an underlying dynamical system. In testing this hypothesis we provide a dynamical systems mechanism to explain how evoked changes in activity affect changes in moment-to-moment variability and functional connectivity. We demonstrate that a simple network model can reproduce emergent statistical phenomena widely described throughout the task-evoked and dynamic functional connectivity literature. Further, our model suggests that evoked activity shifts the system’s attractor dynamics, inducing changes to the moment-to-moment variability and functional connectivity within the network. Together, the proposed mechanisms provide direct links between intrinsic and evoked activity, variability, and functional connectivity under a single dynamical systems framework.