Identification of functional synaptic plasticity from ensemble spiking activities: A nonlinear dynamical modeling approach

摘要

This paper presents a systems identification approach for studying the long-term neural plasticity using natural ensemble spiking activities recorded from behaving animals. It is designed to quantify and explain the non-stationarity in the input-output properties of a brain region. Specifically, we propose a three-step strategy for such a goal. First, a multiple-input, multiple-output (MIMO) nonlinear dynamical model is formulated to estimate and represent the synaptic strength in means of functional connectivity between input and output neurons. Second, this MIMO model is extended to a time-varying form and used to track the non-stationary properties of functional connectivity. Finally, an ensemble synaptic learning rule is identified to explain the input-output non-stationary as the consequence of the past input-output spiking patterns. This framework can be used to study the underlying mechanisms of learning and memory in behaving animals, and may serve as the computational basis for building the next-generation adaptive cortical prostheses.