Topography, Spectral Characteristics, and Extra-to-Intracranial Propagation Pathways of EMG

Objective Intracranial EEG (iEEG) plays an increasingly important role in neuroscientific research and can provide informative control signal for brain-machine interfaces (BMI). While it is clear that electromyographic (EMG) activity of extracranial origin reaches intracranial recordings, the topographic and spectral characteristics of intracranial EMG have been scarcely investigated. It is currently unclear how these characteristics compare to those of physiological brain activity. Little is also known about the exact pathways of extra- to intracranial volume conduction, including the role of craniotomy defects.Methods In 5 epilepsy patients under invasive pre-neurosurgical EEG monitoring, we examined chewing-related effects (ChREs) as a source of intracranial EMG activity and compared those effects with physiological brain activity of 9 patients during several behavioural tasks. These included speech production, finger movements, and music perception. Further, we analyzed the association of craniotomy defects (burr-holes and saw-lines) and the intracranial EMG-effects based on the individual post-operative images.Results ChRE presented with a spatially smooth distribution across almost all intracranial electrodes with the maximum below the temporal muscle. In contrast, the responses of neural origin were spatially more focalized. ChREs were broad-banded and had a higher spectral power and affected higher frequencies than event-related neural activity. ChRE were largely independent of the individual configuration of craniotomy defects. However, we found indications that the silicone sheet, in which electrocorticography (ECoG) electrodes are embedded, attenuates EMG influences, when sufficiently large.Conclusion The present work is the first comprehensive evaluation of topographic and spectral characteristics of EMG effects in iEEG based on a large sample of subjects. It shows that chewing-related EMG can affect iEEG recordings with higher power than typical physiological brain activity, especially in higher spectral frequencies. As the topographic pattern of ChRE is largely independent of the individual position of craniotomy defects, a direct pathway of volume conduction through the intact skull plays an important role for extra- to-intracranial signal propagation. Intracranial EMG activity related to natural behavior should be accounted for in neuroscientific and BMI applications, especially when based on high-frequency iEEG components. A detailed knowledge of EMG properties may help to design both EMG-reducing algorithms and ECoG grids with a high shielding factor.Highlights