Utilizing Slope Entropy as an Effective Index for Wearable EEG-Based Depth of Anesthesia Monitoring

Annu Int Conf IEEE Eng Med Biol Soc. 2024 Jul:2024:1-4. doi: 10.1109/EMBC53108.2024.10782706.

Abstract

Based on prior research indicating a decrease in the spectral slope of electroencephalogram (EEG) during anesthesia induction and an increase during recovery, we propose Slope Entropy (SlopEn), which uniquely emphasizes variations in signal slope, as a new index for monitoring the depth of anesthesia (DoA). The performance of SlopEn is investigated on just a single frontal EEG channel and is compared against other well-known entropy metrics utilized in the field. After filtering the EEG signal, four types of entropy, including SlopEn, are derived from all EEG sub-bands and separately inputted to a regressor for estimating DoA index values. Comparing the results obtained using SlopEn with those from the Sample entropy demonstrates the superiority of the former, achieving a higher correlation coefficient (0.75 vs. 0.63) and a lower median absolute error (4.2 vs. 6.2) between the estimated and reference DoA index values. These findings establish that the SlopEn has the potential to become a valuable index for DoA monitoring using single frontal channel EEG systems.

MeSH terms

  • Algorithms
  • Anesthesia*
  • Electroencephalography* / instrumentation
  • Electroencephalography* / methods
  • Entropy
  • Humans
  • Signal Processing, Computer-Assisted
  • Wearable Electronic Devices*