Background

Identifying the state-related “neural correlates of consciousness” for anesthetics-induced unconsciousness is challenging. Spatiotemporal complexity is a promising tool for investigating consciousness. We hypothesized that spatiotemporal complexity may serve as a state-related but not drug-related EEG indicator during an unconscious state induced by different anesthetic drugs (e.g., propofol and esketamine).

Methods

We recorded EEG from patients with unconsciousness induced by propofol (n=10) and esketamine (n=10). Both conventional microstate parameters and microstate complexity were analyzed. Spatiotemporal complexity was constructed by microstate sequences and complexity measures. Two different EEG microstate complexities were proposed to quantify the randomness (type I) and complexity (type II) of the EEG microstate series during the time course of the general anesthesia.

Results

The coverage and occurrence of Microstate E (pre-frontal pattern) and the duration of Microstate B (right frontal pattern) could distinguish the states of pre-induction wakefulness, unconsciousness, and recovery under both anesthetics. Type I EEG microstate complexity based on mean information gain significantly increased from awake to unconsciousness state (propofol: from mean ± SD of 1.562 ± 0.059 to 1.672±0.023, p < 0.001; esketamine: 1.599±0.051 to 1.687±0.013, p < 0.001), and significantly decreased from unconsciousness to recovery state (propofol: 1.672±0.023 to 1.537±0.058, p < 0.001; esketamine: 1.687±0.013 to 1.608±0.028, p < 0.001) under both anesthetics. In contrast, type II EEG microstate fluctuation complexity significantly decreased in the unconscious state under both drugs (propofol: from 2.291±0.771 to 0.782±0.163, p < 0.001; esketamine: from 1.645±0.417 to 0.647±0.252, p < 0.001), and then increased in the recovery state (propofol: 0.782±0.163 to 2.446±0.723, p < 0.001; esketamine: 0.647±0.252 to 1.459±0.264, p <0.001).

Conclusions

Both type I and type II EEG microstate complexities are drug-independent. Thus, the EEG microstate complexity measures we proposed are promising tools for building state-related neural correlates of consciousness to quantify anesthetic-induced unconsciousness.