Related Experiment Video
Updated: May 5, 2026
Author Spotlight: Investigating Anesthesia-Induced Sleep Pathways and Neuronal Excitability in Mice
Published on: October 11, 2024
The Perioperative Neurocognitive Disorder Prediction Based on AI-Assisted EEG Dynamic Features in Anesthetized Mice
Xinyang Li1, Hui Wang2,3,4, Qingyuan Miao1
1Shanghai Key Laboratory of Anesthesiology and Brain Functional Modulation, Translational Research Institute of Brain and Brain-like Intelligence, Clinical Research Centre for Anesthesiology and Perioperative Medicine, Department of Anesthesiology and Perioperative Medicine, Shanghai Fourth People's Hospital, School of Medicine, Tongji University, Shanghai 200434, China.
Aging increases the risk of postoperative neurocognitive disorders (PND). This study identified specific electroencephalography (EEG) patterns during anesthetic emergence in aged mice that predict PND vulnerability.
Area of Science:
- Neuroscience
- Anesthesiology
- Gerontology
Background:
- Postoperative neurocognitive disorders (PND) are common in elderly surgical patients.
- Aging is a significant risk factor for developing PND.
Purpose of the Study:
- To identify electrophysiological markers for PND prediction.
- To establish a machine learning framework for PND vulnerability using anesthetic electroencephalography (EEG) features in aged mice.
Main Methods:
- Young and aged mice underwent surgery under isoflurane anesthesia with continuous EEG monitoring.
- Neurocognitive function was assessed using standardized behavioral tests.
- A semi-supervised K-means algorithm clustered aged mice into PND and non-PND groups based on young mice.
- Machine learning models were trained to predict PND from EEG features during anesthesia maintenance and emergence.
Main Results:
- Aged mice showed spatial and contextual memory impairments after surgery, with two-thirds classified as PND.
- PND mice exhibited increased delta power and decreased alpha/beta ratios during anesthetic emergence.
- A Multi-layer Perceptron classifier achieved high discriminatory performance for PND prediction (AUC = 0.94).
Conclusions:
- Emergence-related EEG features are associated with postoperative neurocognitive vulnerability in aged mice.
- An exploratory machine learning framework can aid in preclinical PND risk stratification.
- Findings warrant further investigation and validation in human perioperative EEG data.
