Alteration of Temporal Organization of EEG Microstate Sequences During Propofol-induced Loss of Consciousness

• Aim of the study: Assess the alteration of temporal organization of EEG microstate sequences during propofol-induced loss of consciousness
• Methods: Prospective clinical trial. 20 right-handed adult patients, aged between 18 and 40 years, scheduled for elective surgery under general anaesthesia, will be included.

Patients will not receive any preoperative oral anxiolysis. After arrival in the operation theatre and a resting period of 10 minutes, the baseline EEG will be recorded (5 minutes duration). Then, after a three minutes proxygenation period with 100% oxygen, patients will receive an intravenous induction with propofol using the pharmacokinetic model by Schnider et al. The initial cerebral concentration will be 0.5 µg ml-1, which will be increased stepwise by 1.0 µg ml-1 until 2.5 µg ml-1, and then by 0.5 µg ml-1 until loss of consciousness. During the induction procedure, the patient's lungs will be gently ventilated using 100% oxygen through a face mask.

Five minutes after reaching each equilibration of the blood-brain propofol concentration, clinical sedation (using the validated six points Observer Assessment of Alertness/Sedation [OAA/S] scale) will be annotated. Raw EEG, used later for fractal analysis, will be continuously recorded during the procedure. Corresponding OOA/S scores will be recorded on raw EEG. The study ends 10 minutes after the patient has lost consciousness (absence of response to "mild prodding or shaking" corresponding to OAA/S <2).

The fractal analysis of EEG will be performed with a delay after anaesthesia by neuroscientists

• Hypothesis: We hypothesise that the fractal properties of EEG microstates will be modified in parallel with the propofol-induced loss of consciousness.
• EEG and DATA analysis Microstate analysis First, we will determine the maxima of the global field power (GFP). Because topography remains stable around peaks of the GFP, they are the best representatives of the momentary map topography in terms of signal-to-noise ratio.18 All maps marked as GFP peaks (i.e., the voltage values at all electrodes at that time point) will be extracted and submitted to a modified spatial cluster analysis using the atomize-agglomerate hierarchical clustering (AAHC) method19 to identify the most dominant map topographies. The optimal number of template maps will be determined by means of a cross-validation criterion.20 We will then submit the template maps identified in every single subject into a second AAHC cluster analysis to identify the dominant clusters across all subjects. Finally, we will compute a spatial correlation between the templates identified at the group level and those identified for each subject in every run. We will so label each individual map with the group template it best corresponds to, to use the same labels for the subsequent group analysis.