Neuronal and Hemondynamic Responses from Functional MRI Time-Series: A Computational Model

摘要

A computational model to represent neuronal and hemodynamic events in fMRI time-series of human brain activation in sensory and cognitive experiments is presented. The transient and habituated neuronal activities are modeled with exponential functions and vascular coupling with a Gaussian function. An algorithm to estimate model parameters from fMRI time-series is proposed and its accuracy is demonstrated with simulated time-series. This paper highlights the potential application of fMRI for in vivo study of nuronal populations in human brain activations at macroscopic level.