Multi-scale sample entropy of electroencephalography during sevoflurane anesthesia

Area of Science:

• Anesthesiology
• Neuroscience
• Biomedical Engineering

Background:

• Electroencephalogram (EEG) is crucial for assessing depth of anesthesia (DoA).
• Multi-scale EEG analysis offers a more comprehensive understanding of brain dynamics than traditional non-linear methods.
• Modified sample entropy (MSpEn) is a novel approach for analyzing anesthetic EEG series.

Purpose of the Study:

• To evaluate the efficacy of MSpEn in measuring DoA during sevoflurane anesthesia.
• To compare MSpEn with established EEG indices: approximate entropy (ApEn), response entropy (RE), and state entropy (SE).
• To investigate the influence of scale number on MSpEn performance.

Main Methods:

• Analysis of EEG data from 19 adult subjects undergoing sevoflurane anesthesia.
• Computation of MSpEn, ApEn, RE, and SE for EEG series.
• Utilized pharmacokinetic/pharmacodynamic modeling and prediction probability (Pk) for performance assessment.
• Employed linear regression to determine the optimal number of scales for MSpEn.

Main Results:

• MSpEn demonstrated the highest prediction probability (Pk = 0.83 ± 0.05) and coefficient of determination (R² = 0.87 ± 0.21) in tracking sevoflurane concentration.
• MSpEn outperformed ApEn, RE, and SE in correlating with anesthetic effect.
• Scales 1, 3, and 5 contributed most significantly to MSpEn's performance.

Conclusions:

• MSpEn is a more effective method than ApEn, RE, and SE for monitoring sevoflurane concentration changes.
• The findings suggest MSpEn's potential for clinical application in DoA monitoring.
• Further research into MSpEn for clinical DoA assessment is warranted.