Time-resolved phase-amplitude coupling (PAC) in TMS-induced epileptiform discharges (ED): a multichannel inter-regional analysis pipeline

Epileptiform discharges (EDs) perturb neural communication on a millisecond-to-second scale, requiring EEG analyses with fine temporal resolution. Phase-Amplitude Coupling (PAC) occurs when the instantaneous amplitude (amplitude envelope) of a fast brain oscillation (e.g., γ) is statistically dependent on the phase of a slower oscillation (e.g., δ or θ). Non-invasive perturbation of neural circuits via Transcranial Magnetic Stimulation (TMS) allows probing cross-frequency interactions like PAC during induced EDs. We present a MATLAB pipeline for extracting time-resolved PAC from multichannel TMS-EEG recordings of epilepsy patients; the pipeline is applied to data from 4 patients with 11 ED events using sliding windows at sub-second resolution. The analysis provides regional mapping of PAC dynamics throughout ED events, and considers multiple complementary PAC metrics (including information-theoretic estimators) to support cross-metric interpretation. Across patients and metrics, the most robust finding was a transient, ED-locked surge of low-frequency phase ~ high-frequency amplitude coupling, predominantly in the δ ~ γ and θ ~ γ bands. This formed a pan-regional, anterior-central dominated network state that clearly distinguished the ED state from pre-ED and post-ED epochs. Localized posterior high-γ effects around EDs and persistent temporal lobe PAC signatures in some patients pointed to additional, region-specific mechanisms. The information-theoretic MIPAC estimators further revealed a systematic β-band desynchronization during ED that amplitude-sensitive metrics largely missed. Overall, the proposed pipeline provides statistically controlled mapping of transient inter-regional PAC dynamics in TMS-EEG recordings from epilepsy patients before, during, and after EDs. This offers a general framework for studying PAC changes in EDs, as well as brain responses to stimulation and perturbation across diverse neurological conditions beyond epilepsy, thereby enhancing our understanding of epileptogenesis and informing future neuromodulation strategies.