Recording Brain Electromagnetic Activity During the Administration of the Gaseous Anesthetic Agents Xenon and Nitrous Oxide in Healthy Volunteers

The purpose of this experiment is to characterize changes in brain electromagnetic activity in response to the administration of the two gaseous dissociative agents xenon and nitrous oxide to healthy participants in a two-way repeated measures crossover design. Due to the complexity of the experiment, it is important that the majority of equipment setup occurs before participant arrival. The most important of these is the configuration of the EEG system in the magnetically shielded room or MSR.

The EEG amplifier is placed in the MSR as far as practical away from the participant, and needs to be configured to run on DC power using either a built-in battery or an external shielded battery box. Fiber optic coupling from the amplifier to the data-logging computer should be checked to be operational. Finally, a video camera in the MSR should be configured such that its field of view clearly captures the participant's face and upper thorax such that the physician anesthesiologist can clearly monitor the respiratory and airway status of the participant.

In order that consistent and comparable data is acquired, participants are first interviewed to confirm their eligibility for the study. Participants should be healthy males of 20 to 40 years of age with no psychiatric or neurological conditions, no recent history of drug or alcohol use, and have no respiratory or cardiac conditions. Participants should arrive clean-shaven on the day of the experiment, and having fasted for at least six hours.

To ensure high quality EEG recordings, it is recommended that the subject's hair has been recently washed without conditioner, and that their scalp has been firmly brushed to remove any dried or dead skin. An MEG-compatible high-density EEG cap is applied with the appropriate use of conductive gel. Head position indicator coils are then firmly attached at predefined locations on the EEG cap.

All auxiliary biopotential electrodes are then applied at the designated locations on the subject's face and body. For subsequent analysis, it is important that all electrodes and HPI coil positions are accurately registered along with digitization of the outlines of the orbits, nasal ridge and cheeks for subsequent coregistration with structural MRI. Two electrodes are placed submentally to record resting muscle tone.

Electrooculogram or EOG recordings are made by placing two further electrodes above and below the eye. And finally, a three-lead ECG derivation is configured by electrodes on the left and right wrist flexor surfaces. EEG electrode impedances should then be verified as being below five kiloohms in order to yield high-fidelity recordings.

The duration of the experiment as well as the sedative effects of the administered gases makes minimization of head movement vital. A memory foam head brace is applied in order to reduce lateral head movement, and a jute harness is tied around the body to minimize any vertical movement. Eyes-closed resting state and baseline auditory continuous performance task recordings are now performed.

It is of vital importance that the administration of anesthetic gases is supervised by a physician anesthesiologist at all times. As well as patient monitoring, the role of the physician anesthesiologist includes ensuring the adequate configuration of the anesthetic gas delivery unit, specialized in this case to deliver and monitor xenon. The physician anesthesiologist will test the anesthesia machine, check that the mask fits comfortably and does not leak, adjust gas flows, insert a cannula, administer antiemetics, as well as commencing denitrogenation with 100%oxygen with the participant seated in the MSR before the final post-antiemetic baseline data collection and the start of the gas administration protocol.

At this stage the participant will be comfortably seated in the MSR having been denitrogenated and having been administered dexamethasone and ondansetron to minimize the risk of any adverse nausea or vomiting due to gas administration. Before the MSR door is closed, the anesthetic nurse or other suitably trained medical professional is seated next to the participant. A final check is made that the blood pressure cuff is connected, the button response boxes are comfortably positioned, and the pulse oximeter on the toe or finger is securely fastened.

The protocol then proceeds to deliver four Equi-MAC awake progressively increasing stepwise levels of xenon gas, or three stepwise nitrous oxide levels over a period of 15 minutes each in order to assure blood-brain-gas equilibration. Each level should contain a five-minute steady state when the desired gas concentration is reached and should be followed by a 10-minute washout period. During all levels, oxygen, carbon dioxide, and xenon or nitrous oxide levels are continuously monitored and electronically logged.

At the end of each level, participants are briefly assessed before progressing to higher levels. Because at the highest xenon level, in which loss of responsiveness is achieved, airway obstruction can readily and unexpectedly occur, and subjects need to be closely monitored and continuously assessed by qualified personnel, both inside and outside the MSR for all necessary action to be taken immediately. At the conclusion of the experiment, the physician anesthesiologist will enter the MSR and debrief with the patient.

The participant is then brought outside the MSR where the apparatus is removed and they are administered standard postoperative anesthetic care. Before the participant leaves, ensure they are safe to travel home with a relative or friend. In the next 24 hours, the participant will be required to fill in the questionnaires regarding the procedure and any after effects.

This video has been primarily prepared to illustrate the methods involved in delivering gaseous anesthesia while recording simultaneous EEG and MEG data, therefore, only very basic data analysis and representative results are presented to give the viewer a clear idea of what the recorded data should look like. The basic results illustrate examples of head movement, gas concentration timecourse and auditory task accuracy;filtered artifact-free EEG and MEG data;EEG and MEG spectra and power topography;and auditory evoked responses for both baseline and during gas administration. To create the basic results, the following pipeline is employed using a combination of data analysis software platforms and toolboxes.

The raw MEG and EEG data is stored as fif and cnt file formats respectively. Head movements are computed using the MaxFilter software and a script written by the authors, both used to convert head movements from the cortonian space into Cartesian coordinates. Auditory task accuracy and reaction time are then computed and visualized, time-aligned with the head movement and gas concentration time series, in order to assess data quality before further processing the EEG and MEG data.

Time alignment of the data is based on recorded trigger times. Typically, head movement will be minimal during baseline conditions, while more movement tends to occur at the higher gas levels associated with more dissociated or sedated states. This can be seen here for example in the 42%xenon recording where the top figure shows head movement, and the bottom, gas concentration and auditory task accuracy.

Note that loss of responsiveness occurs between 7.5 and 12.5 minutes. Absolute head movement of less than five millimeters is considered acceptable for source imaging purposes. Data segments with movement larger than five millimeters are ignored.

Preprocessing of the MEG and EEG data initially involves selection of bad channels. This is followed by the use of the tSSS algorithm to remove artifacts from the MEG recording due to external sources. The rest of the processing pipeline involves the use of data analysis software platforms and associated toolboxes for analysis.

Band-pass and notch filtering is applied to both the MEG and EEG data to focus analysis on the frequency band of interest. Time periods in the data containing artifacts are rejected based on visual inspection, in addition to removing periods involving excessive head movement changes. Filtered artifact-free MEG and EEG data can then be visualized and related to anesthetic-induced changes.

Amplitude spectra for each sensor and spectral band power topographies across all sensors are then computed for both MEG and EEG data. Visualization of the amplitude spectra at different sensor locations facilitates observation of the effects of different anesthetic levels on signal spectra. Visualization of sensor level power topography in different frequency bands, in this case the alpha band, enables the observation of the effects of different anesthetic levels on the spacial pattern of activity.

The filtered artifact-free MEG and EEG data are then epoched using triggers linked to the time of auditory tone delivery. Average auditory evoked responses are then constructed through time on sample averaging. Visualization of sensor level auditory evoked responses allows for the characterization of the effects of anesthetics on the brain's responsiveness.

In particular, awakening of the initial evoked response and the disappearance of the later evoked response peaks results from high gas concentrations. There are many other potential interesting analysis steps that can be performed subsequent to this basic analysis. Given that there are many options presented in the literature, we do not cover them in detail here.

Such options include MEG and EEG source imaging using beamforming or minimum-norm estimation to gain a better understanding of the relationship between brain activity and brain anatomy. Sensor or source-level functional connectivity analysis can unravel changes in brain networks linked to anesthetic action. There is also the option to either focus analysis on the sensor or source-level auditory evoked responses, or to look at continuous data independent of the auditory stimulus in order to identify any anesthetic-induced changes in global brain dynamics.

The macroscale mechanisms of anesthetic-induced reductions in consciousness are not well understood. Evidence from literature has strongly implicated a breakdown in functional connectivity underlying anesthetic-induced reductions in consciousness. Although anesthetics have a diverse range of molecular targets of action and different macroscopic effects on the EEG and MEG, it is still not resolved if there exists a universal or common macroscale mechanism across all anesthetics.

Here we have outlined a protocol for performing gaseous anesthesia using the putative NMDA receptor antagonists xenon and nitrous oxide while simultaneously recording MEG and EEG data. This method will facilitate the study of different gaseous and volatile agents on simultaneously recorded MEG and EEG data, and thus the search for universal mechanisms of anesthetic-induced reductions in consciousness.