Neuronal Models for EEG–fMRI Integration

Abstract

Human brain activity can be measured in many ways, providing different views into its functions. Common measurements of human brain function, such as functional MRI (fMRI) and electroencephalography (EEG), are often integrated to get the best spatial and temporal resolution, but these measurements frequently show differences. This chapter therefore considers how the pooling of signals across populations of neurons affects these measurements. We consider a modeling framework in which fMRI and field potential signals integrate across transmembrane currents in a population of neurons, because synaptic activity is thought to be the largest drive of the fMRI signal and transmembrane potentials are the biggest source of field potentials. We formulate computations that provide simplified abstractions that inform how levels of synaptic activity or synchrony across neurons contribute to fMRI or electrophysiological signals and how to interpret deviations or similarities between measurements. The modeling framework and computations highlight that the level of activity in a neuronal population influences each measurement, but synchrony only has a large effect on field potentials and not on the fMRI signal. An application to data from human visual cortex explains why certain signals correlate and others do not. Advancing the fundamental understanding of how different measurements integrate neuronal activity will be important to combine fMRI and EEG measurements to better understand human brain function.