Consciousness is more complex than simply labeling states like deep sleep and anesthesia as "unconscious." Investigations show that behavioral signs and subjective reports can be unreliable, challenging traditional beliefs. For instance, individuals in presumed unconscious states may still report experiences, suggesting that 30-50% of patients under general anesthesia retain some awareness. Reevaluating these assumptions could enhance scientific clarity and improve clinical practices, fostering a deeper understanding of consciousness and its disorders. A nuanced approach is essential for future advancements in the field.
In a follow-up EEG sleep study, brain signal complexity (Lempel-Ziv complexity, LZC) decreased progressively from wakefulness into deeper non-REM sleep. However, within NREM2 sleep, there was no significant difference in LZC between dream and non-dream awakenings, and no correlation between LZC and subjective ratings of dream vividness, diversity, or perceptual quality. The authors failed to reproduce their earlier finding that posterior LZC increased with more perceptual dream experiences. This raises doubts about whether EEG LZC is a reliable marker of richness of experience within the same sleep stage.
Signal diversity in EEG recordings, measured by Lempel-Ziv complexity and other metrics, decreases with deeper non-REM sleep stages, consistent with theories linking consciousness to complex cortical dynamics. However, signal diversity did not significantly differ between dreaming and non-dreaming periods within the same sleep stage. A positive correlation was found between Lempel-Ziv complexity over the posterior cortex and the thought-perceptual quality of dream contents, suggesting that specific aspects of dream experience may relate to local cortical signal diversity.
A systematic review of 255 EEG-based measures of consciousness found that signal diversity and event-related potential measures are the most consistent for distinguishing conscious from unconscious brain states. Spectral entropy, Lempel Ziv complexity, and spectral edge frequency emerged as the most practical, consistent, and reproducible measures. However, because most studies relied on behavioral assessments rather than subjective reports, the true presence or absence of phenomenological experience remains uncertain, limiting the conclusions that can be drawn. The review provides detailed categorizations to guide future research.