Fig. 1: Whole-cortex modeling and parameter estimation framework.

a The schematic of the whole-cortex model fitted to source-level MEG data is shown. Each MEG source time series is modeled using an NMM, which represents a node in the whole-cortex model. The connections between the NMMs are indicated by purple lines. b 78 ROI MEG source time series from an example subject during Xenon-induced anesthesia. c, d ACPT accuracy time series and Xenon concentration levels for the subject. LC and HC represent the estimated onset of low- and high- level consciousness, respectively. e Time-series estimation of the strength of inter-regional connectivity between NMMs. f Time-series estimation of local model states and connectivity parameters was performed for an NMM fitted to the right precuneus of the subject. g Time courses of averaged local and inter-regional connection strength across all subjects, with the 95% confidence interval shown as gray shaded bands.