Lower EEG alpha band power is currently extensively studied as a potential biomarker of brain vulnerability. No previous study has evaluated whether there was a difference in EEG power spectrum ranges between propofol and sevoflurane, the two main anesthetic agents for maintenance of general anesthesia, in order to define a cut-off value for the prediction of a higher risk of postoperative delirium.

• Primary objective: comparison of EEG alpha power (dB) during maintenance of anesthesia with propofol and sevoflurane using data available in the NeuroSense® depth-of-anesthesia monitoring
• Secondary objectives: comparison of other EEG frequency bands during maintenance of anesthesia with propofol and sevoflurane using data from the NeuroSense® depth-of-anesthesia monitoring

Since the development and commercialization of quantitative EEG monitors, which compute an index to assess the depth of sedation, the interest of anesthesiologists for neurophysiology has increased considerably. Indeed, frontal EEG represents an accessible, non-invasive technique to monitor the brain, the target-organ of our hypnotic drugs.

Anesthesia interferes with the physiological oscillations generated by the brain, and these disturbances impact the EEG. For instance, common hypnotic agents used for the maintenance of general anesthesia, intravenous propofol or inhaled sevoflurane, provide a typical pattern on the raw EEG related to their main mechanism of action: predominant simultaneous frontal slow/delta (0.1-4 Hz) and alpha (8-12 Hz) oscillations.

Intraoperative alpha oscillations are currently being extensively studied, as they might be a potential biomarker of brain vulnerability, which could help in predicting postoperative delirium, one of the most prevalent neurological complication after surgery. Recently, in a prospective observational cohort study including 220 patients, we showed that adult patients undergoing elective cardiac surgery under sevoflurane anesthesia who had lower intraoperative frontal EEG alpha band power were at higher risk of postoperative delirium. Lower intraoperative frontal EEG alpha band power was also correlated with poorer preoperative cognitive scores in our clinical series, which confirms previously published observations.

However, some practical questions still need to be addressed before this electroencephalographical marker can be integrated into a risk-prediction model for postoperative delirium. More specifically, we are not currently able to suggest a "universal" cut-off value for intraoperative frontal alpha-band power under general anesthesia that could be useful for a daily clinical practice. For example, an important issue is that the definition of the cut-off value will highly depend on the choice of anesthetic drugs. Although propofol and volatile anesthetics, such as sevoflurane, share similar neurophysiological characteristics, their respective EEG patterns and spectrograms are not identical (increased power in the thêta-band with volatile anesthetics). Otherwise, both propofol and sevoflurane EEG patters become completely different when they are used in combination with other anesthetic agents, such as benzodiazepines, ketamine or dexmedetomodine. These combinations are frequently used in multimodal anesthesia practice, and can deeply impact EEG frequency band power values. Consequently, at the moment, "low" alpha band power remains rather an observation on our patients' EEG spectrogram throughout surgery than a quantifiable value.

The main objective of this study is a first step towards to definition of a cut-off value for EEG frontal alpha band power to predict postoperative delirium. We aim to determine whether there is a significant difference in EEG power spectrum ranges between propofol and sevoflurane, the two main anesthetic maintenance agents. More precisely, it would require "simple", non-balanced, maintenance of general anesthesia using either propofol or sevoflurane, with the avoidance of any other hypnotic drugs. This type of maintenance is frequently used for minor interventions requiring general anesthesia.

Inclusion Criteria: Adult patients undergoing elective electrophysiological interventions, such as atrial fibrillation ablation procedures, under general anesthesia

• Capable to give informed consent
• French-speaking patients

Exclusion Criteria:

• Need for a premedication with benzodiazepine before anesthesia
• Reduced left ventricular ejection fraction (< 50%) or underlying cardiac pathology requiring a more balanced anesthesia (administration of benzodiazepine and/or ketamine during anesthesia)
• Preoperative MoCA score < 26/30 (for patients > ou = 55 years old)
• Patients treated with anti-epileptic drugs
• Patients treated with benzodiazepines (ex: for anxiolysis or sleep disorders)
• Emergencies
• Morbid obesity (BMI > 40kg/m2)
• Background of severe postoperative nausea/vomit