Decoding personalized motor cortical excitability states from human electroencephalography
OBJECTIVE: Brain state-dependent transcranial magnetic stimulation (TMS) requires real-time identification of cortical excitability states. Here, we aimed to identify individualized, subject-specific motor cortex (M1) excitability states from whole-scalp electroencephalography (EEG) signals. METHODS: We analyzed a pre-existing dataset that delivered 600 single TMS pulses to the right M1 during EEG and electromyography (EMG) recordings. Subject-specific multivariate pattern classification was used to discriminate between brain states during which TMS elicited small or large motor-evoked potentials (MEPs). RESULTS: Classifiers trained at the individual subject level successfully discriminated between low and high M1 excitability states. MEPs elicited during classifier-predicted high excitability states were significantly larger than those elicited during classifier-predicted low excitability states. Classifiers trained on subject-specific data obtained immediately before TMS delivery performed better than classifiers trained on data from earlier time points, and subject-specific classifiers generalized weakly but significantly across subjects. CONCLUSION: Decoding individualized M1 excitability states from whole-brain EEG activity is feasible and robust. SIGNIFICANCE: Deploying subject-specific classifiers during brain state-dependent TMS may enable effective, fully individualized neuromodulation in the future.