EEG phase entropy distinguishes conscious from unconscious brain states with high accuracy
Researchers used permutation entropy on EEG phase dynamics to classify consciousness levels, but ADHD detection remains elusive.
In this study, published in Chaos, Solitons and Fractals and available on arXiv, researchers from Korea and the US applied permutation entropy analysis to EEG phase dynamics, focusing on the principal mode reflecting anterior-posterior information flow. They analyzed two datasets: one from a general anesthesia protocol (conscious vs. unconscious) and one from resting-state EEG of healthy controls and individuals with inattentive-type ADHD under eyes-open and eyes-closed conditions.
The results showed that permutation entropy distributions reliably distinguish conscious from unconscious states in the anesthesia dataset, and eyes-open from eyes-closed in the resting-state dataset. However, the method failed to achieve clear separability between control and ADHD groups, indicating that additional features — such as original time-series values — may be necessary for detecting inattentive-type ADHD. The study demonstrates both the power and limitations of ordinal pattern analysis for brain state classification.
- Permutation entropy of EEG phase dynamics distinguishes conscious from unconscious states with high reliability in general anesthesia data.
- Eyes-open and eyes-closed resting-state conditions are separable using the same method.
- Classification between healthy controls and inattentive-type ADHD patients shows no clear separability with permutation entropy alone.
Why It Matters
This work could improve anesthesia monitoring and deepen understanding of consciousness, while highlighting EEG analysis limits for ADHD diagnosis.