Speaker
Description
Current commercial EEG anesthesia monitoring systems put older patients at risk of overdose \cite{ni2019,obert2021influence}. Such monitors rely on frequency-domain EEG analysis, which assumes quasi-periodic stationary signals \cite{cohen2014}, but omitting phase data causes information loss. As EEG waveforms are often asymmetric \cite{cole2017}, an overlooked feature in the context of anesthesia may be the asymmetry of EEG oscillatory activity in the time-domain.
Waveform asymmetry measures were assessed across frequency bands in two independent datasets with respect to age and anesthetic compound. Additionally, the Hilbert envelope of EEG bands was extracted and characterized using the same measures, zero-crossing variability, and sinusoidal parametric model fitting.
Sharpness and steepness asymmetry decreased with patient age in a frequency-dependent manner, strongest in the delta band. Sevoflurane and desflurane showed stronger concentration-dependent effects than propofol. Envelope analysis revealed consistent sharpness asymmetry across all frequency bands. High zero-crossing variability and limited parametric model fit suggest that simple sinusoidal models may not fully capture the envelope.
Building on these findings, we aim to investigate whether asymmetric waveforms can be described by computational models to develop more physiologically accurate EEG representations.
Bibliography
@article{ni2019,
title={Paradox of age: older patients receive higher age-adjusted minimum alveolar concentration fractions of volatile anaesthetics yet display higher bispectral index values},
author={Ni, Katherine and Cooter, Mary and Gupta, Dhanesh K and Thomas, Jake and Hopkins, Thomas J and Miller, Timothy E and James, Michael L and Kertai, Miklos D and Berger, Miles},
journal={British journal of anaesthesia},
volume={123},
number={3},
pages={288--297},
year={2019},
publisher={Elsevier}
}
@book{cohen2014,
title = {Analyzing Neural Time Series Data: Theory and Practice},
author = {Cohen, Mike X.},
year = {2014},
publisher = {MIT Press},
address = {Cambridge, MA},
isbn = {9780262019873},
doi = {10.7551/mitpress/9609.001.0001}
}
@article {cole2017,
author = {Cole, S.R. and Voytek, B.},
title = {Brain oscillations and the importance of waveform shape},
journal = {Trends in Cognitive Sciences},
volume = {21},
number = {2},
pages = {137--149},
year = {2017}
}
@article{obert2021influence,
title={The influence of age on EEG-based anaesthesia indices},
author={Obert, David P and Schweizer, Catrin and Zinn, Sebastian and Kratzer, Stephan and Hight, Darren and Sleigh, Jamie and Schneider, Gerhard and Garc{\'\i}a, Paul S and Kreuzer, Matthias},
journal={Journal of clinical anesthesia},
volume={73},
pages={110325},
year={2021},
publisher={Elsevier}
}
