Electroencephalographic signals during anesthesia recorded from surface and depth electrodes

PURPOSE

Anesthesiologists have increasingly started to use EEG-based indexes to estimate the level and type of unconsciousness. However, the physiology and biophysics are poorly understood in anesthesiological literature.

METHODS

EEG was recorded from electrodes on the surface of head, including scalp, as well as DBS (deep brain stimulation) electrodes implanted deep in the brain. Mathematical modeling with a realistic head model was performed to create illustrative images of the sensitivity of electrode montages.

RESULTS

EEG pattern of anesthesia, burst-suppression, is recordable outside of scalp area as well in the depth of brain because the EEG current loops produce recordable voltage gradients in the whole head. The typical electrodes used in anesthesia monitoring are most sensitive to basal surface of frontal lobes as well as frontal and mesial parts of temporal lobes.

CONCLUSIONS

EEG currents create closed-loops, which flow from the surface of the cortex and then return to the inside of the hemispheres. In the case of widespread synchronous activity like physiological sleep or anesthesia, the currents recorded with surface and depth electrodes return through the base of brain and skull.

Proof of concept evaluation of the electroencephalophone as a discriminator between wakefulness and general anaesthesia

BACKGROUND

Depth of anaesthesia (DOA) monitors based on the electroencephalogram (EEG) are commonly used in anaesthetic practice. Their technology relies on mathematical analysis of the EEG waveform, generally resulting in a number which corresponds to anaesthetic depth. We have created a novel method of interpreting the EEG, which retains its underlying complexity. This method consists of turning the EEG into a sound: the electroencephalophone (EEP).

METHODS

In a pilot study, we recorded awake and anaesthetized EEGs from six patients. We transformed each EEG into an audio signal using a ring buffer with a write frequency of 1 kHz and a read frequency of 48 kHz, thus elevating all output frequencies by a factor of 48. In essence, the listener hears the previous 12 s of EEG data compressed into 250 ms, updated every 250 ms. From these data, we generated a bank of 5 s audio clips, which were then used to train and test a sample of 23 anaesthetists.

RESULTS

After training, 21 of the 23 anaesthetists were able to use the EEP to correctly identify the conscious state of >5 of 10 randomly selected patients (P<0.001). The median score was 8 out of 10, with an inter-quartile range of 7-9.

CONCLUSIONS

The EEP shows promise as a DOA monitor. However, extensive validation would be required in a variety of clinical settings before it could be accepted into mainstream clinical practice.