Quantitative evaluation for the wakefulness state using complexity-based decision threshold value in EEG signals.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Prolonged electroencephalography (EEG) measurements can induce drowsiness, complicating EEG signal interpretation.
• Accurate assessment of subject wakefulness is crucial for reliable EEG data analysis.
• Existing spectral-based indices may not sufficiently capture subtle changes in brain activity related to wakefulness states.

Purpose of the Study:

• To develop and evaluate a new quantitative index for assessing subject wakefulness (fully awake vs. drowsy).
• To utilize a complexity-based approach, specifically Approximate Entropy (ApEn), for wakefulness evaluation.
• To compare the efficacy of the proposed index against conventional spectral-based indices.

Main Methods:

• Developed a novel complexity-based index using Approximate Entropy (ApEn) with optimized parameter values.
• Evaluated the index using occipital-alpha rhythm during eye closure in 45 healthy adult subjects.
• Compared the performance of the new index against relative delta power (R.δ), relative theta power (R.θ), and power ratios (Pθ/α, Pθ/β).

Main Results:

• The developed ApEn-based index effectively distinguished between fully awake and drowsy states.
• The proposed index demonstrated superior performance compared to R.δ, R.θ, Pθ/α, and Pθ/β.
• Superiority percentages were 10% for R.δ, 5.5% for R.θ, 8.9% for Pθ/α, and 24.4% for Pθ/β.

Conclusions:

• A new complexity-based index using ApEn offers a superior method for quantitative evaluation of subject wakefulness during EEG recordings.
• This index provides a more sensitive measure of wakefulness compared to traditional spectral analysis techniques.
• The findings suggest improved reliability in EEG interpretation by incorporating this advanced wakefulness assessment tool.