To this date, no monitoring device can reliably detect episodes of accidental awakening during general anesthesia. One promising approach is the use of electroencephalography (EEG) to detect movement attempts via a brain-computer interface (BCI). Previous work has shown that combining a BCI with painless median nerve stimulation can detect cerebral motor activity under light propofol sedation. However, clinical data are still lacking regarding the persistence or otherwise of a cerebral motor response (neural synchronization, or ERS) induced by this stimulation during general anesthesia.
In this new study, the aim is to simultaneously record the EEG centered on the motor cortex and the signal from the SedLine Patient State Index (PSI) to better characterize the evolution of cerebral motor activity before, during general anesthesia, and up to awakening. This approach will allow us to explore the complementarity of the two signals for future automated detection of residual states of consciousness during surgery.
Preliminary data from a previous protocol (STIM-MOTANA) allowed to develop an EEG classification algorithm based on Riemannian geometry, capable of inferring a patient's state of consciousness from cortical responses induced by median nerve stimulation.
The objective of this new study is also to compare the sensitivity of this algorithm with that of PSI, in order to assess its potential as a complementary - or even alternative - indicator of the level of consciousness under general anesthesia.
Investigators hypothesize that it is possible to detect, using EEG, specific brain signatures related to median nerve stimulation, including during propofol-induced general anesthesia. Specifically, neuronal desynchronization and resynchronization phases (ERD/ERS), well characterized in wakefulness or light sedation, could partially persist at higher propofol concentrations. Parallel PSI recording will allow analyzing to what extent these EEG modulations are related to classically measured levels of consciousness, and to validate the relevance of the new algorithm based on Riemannian geometry as a tool for detecting intra-operative arousals.
Right-handed patient
Scheduled for surgery using intravenous anesthesia with propofol concentration
Allergy to propofol, soy, or peanuts.
BMI < 20 or > 30
Pregnant or breastfeeding women
Medical or surgical history that may interfere with median nerve stimulation or EEG signal acquisition (e.g., diabetes, polyneuropathy, central neurodegenerative disease, epilepsy, brain surgery)