Inferring physiological mechanisms in the human cortex by fitting a neural population model to EEG data

This thesis explores how physiological mechanisms in the human cortex can be inferred by fitting a neural population model to electroencephalogram data. We focus on two well-known but poorly understood features of the electroencephalogram, alpha oscillations and alpha-blocking. By fitting the model to electroencephalogram data from 82 individuals we distill the physiological drivers for each phenomenon that are common across all subjects, despite the considerable between-individual variation in electroencephalogram characteristics. We show how the specific mechanisms underlying the two phenomena can be identified by a posteriori fits to electroencephalogram data rather than by a priori hypothesis.