Corticothalamic dynamics during postictal recovery of self-orientation after electroconvulsive therapy

The mechanisms underlying recovery of consciousness after its transient loss remain incompletely understood. Electroconvulsive therapy (ECT) induced generalized seizures disrupt responsiveness and orientation—key functional dimensions related to consciousness—and are followed by a stereotyped postictal state of temporary unresponsiveness, offering a unique model to study the dynamics of connected, report-capable cognition in humans. We performed a post hoc analysis of prospectively collected data from a randomized controlled trial, comprising 345 continuous postictal electroencephalography (EEG) recordings of up to 1 h from 33 patients (median age 53 years, interquartile range 21.3 years; n = 19 [56%] female) undergoing a course of ECT. Using a corticothalamic mean-field model, we estimated cortical (X), corticothalamic loop (Y) and intrathalamic (Z) gain parameters that quantify the responsiveness of cortical and thalamic populations to synaptic input. We examined whether recovery of self-orientation, indexed by the time to reorientation in person, was associated with specific parameter regimes. Immediately after seizure termination, cortical gain was elevated, the corticothalamic loop gain was negative and intrathalamic gain was near zero—suggesting that the thalamus exerted a suppressive effect on the cortex while exhibiting minimal intrinsic activity. Bayesian mixed-effects models showed that during postictal recovery cortical gain × decreased (β = −0.11, CrI₉₅ = [−0.16, −0.07]), while corticothalamic loop gain Y (β = 0.05, CrI₉₅ = [0.02, 0.07]) as well as intrathalamic gain Z (β = 0.05, CrI₉₅ = [0.02, 0.08]) increased, reflecting progressive restoration of thalamic excitatory drive. Across patients and ECT sessions, recovery of self-orientation occurred when model parameters approached towards a characteristic regime, i.e. β_X = 0.81 (CrI₉₅ = [0.74, 0.88]), β_Y = −0.11 (CrI₉₅ = [−0.16, −0.07]) and β_Z = 0.01 (CrI95 = [−0.04, 0.05]), for cortical, corticothalamic loop and intrathalamic gains, respectively. No associations were found between model parameters at which self-orientation was regained and reorientation time, showing relatively small inter-subject variability (sd_X = 0.04, sd_Y = 0.03 and sd_Z = 0.02), suggesting that these values were consistent across patients. To conclude, these findings suggest that restoration of effective corticothalamic coupling represents a critical regime for the re-emergence of self-orientation and reportable responsiveness following postictal unresponsiveness. Our results provide time-resolved modelling of human thalamocortical circuit dynamics during postictal recovery, offering mechanistic insight into how large-scale neural interactions reorganize to restore self-orientation during the postictal state.