Abstract:Background: Proper maintenance of hypnosis is crucial for ensuring the safety of patients undergoing surgery. Accordingly, indicators, such as the Bispectral index (BIS), have been developed to monitor hypnotic levels. However, the black-box nature of the algorithm coupled with the hardware makes it challenging to understand the underlying mechanisms of the algorithms and integrate them with other monitoring systems, thereby limiting their use. Objective: We propose an interpretable deep learning model that forecasts BIS values 25 s in advance using 30 s electroencephalogram (EEG) data. Material and methods: The proposed model utilized EEG data as a predictor, which is then decomposed into amplitude and phase components using fast Fourier Transform. An attention mechanism was applied to interpret the importance of these components in predicting BIS. The predictability of the model was evaluated on both regression and binary classification tasks, where the former involved predicting a continuous BIS value, and the latter involved classifying a dichotomous status at a BIS value of 60. To evaluate the interpretability of the model, we analyzed the attention values expressed in the amplitude and phase components according to five ranges of BIS values. The proposed model was trained and evaluated using datasets collected from two separate medical institutions. Results and conclusion: The proposed model achieved excellent performance on both the internal and external validation datasets. The model achieved a root-mean-square error of 6.614 for the regression task, and an area under the receiver operating characteristic curve of 0.937 for the binary classification task. Interpretability analysis provided insight into the relationship between EEG frequency components and BIS values. Specifically, the attention mechanism revealed that higher BIS values were associated with increased amplitude attention values in high-frequency bands and increased phase attention values in various frequency bands. This finding is expected to facilitate a more profound understanding of the BIS prediction mechanism, thereby contributing to the advancement of anesthesia technologies.

An introduction to bispectral analysis for the electroencephalogram

A Method to Study Global Spatial Patterns Related to Sensory Perception in Scalp EEG.

The Bispectrum and Its Relationship to Phase-Amplitude Coupling

EEG Signal Processing: Theory and Applications

Monitoring of Anesthesia by Bispectral Analysis of EEG Signals

A Reliable Method to Estimate the Bispectral Index Value Using a Single Frontal EEG Channel for Intra and Inter Subject Variability.

Development of a Bispectral index score prediction model based on an interpretable deep learning algorithm

The applied principles of EEG analysis methods in neuroscience and clinical neurology

On the blind source separation of human electroencephalogram by approximate joint diagonalization of second order statistics

Electroencephalography Signal Processing: A Comprehensive Review and Analysis of Methods and Techniques

Sample entropy analysis of EEG signals via artificial neural networks to model patients' consciousness level based on anesthesiologists experience

A neurophysiological basis for aperiodic EEG and the background spectral trend

Bispectrum Analysis of Noninvasive EEG Signals Discriminates Complex and Natural Grasp Types

Characterization of Eeg under Anesthesia and A Comparative Study with Bis

Spectral information of EEG signals with respect to epilepsy classification

EEG Signal Analysis: A Survey

Evaluation of Ischemic States Using Bispectrum Parameters of Eeg and Neural Networks

Epileptic Seizure Detection and Prediction in EEGs Using Power Spectra Density Parameterization

BSD: a Bayesian framework for parametric models of neural spectra

Prediction of Epileptic Seizures Using Bispectrum Analysis of Electroencephalograms and Artificial Neural Network

Advances in Quantitative Electroencephalogram Analysis Methods