Optimizing Motor Intention Detection With Deep Learning: Towards Management of Intraoperative Awareness.

OBJECTIVE: This article shows the interest in deep learning techniques to detect motor imagery (MI) from raw electroencephalographic (EEG) signals when a functional electrical stimulation is added or not. Impacts of electrode montages and bandwidth are also reported. The perspective of this work is to improve the detection of intraoperative awareness during general anesthesia. METHODS: Various architectures of EEGNet were investigated to optimize MI detection. They have been compared to the state-of-the-art classifiers in Brain-Computer Interfaces (based on Riemannian geometry, linear discriminant analysis), and other deep learning architectures (deep convolution network, shallow convolutional network). EEG data were measured from 22 participants performing motor imagery with and without median nerve stimulation. RESULTS: The proposed architecture of EEGNet reaches the best classification accuracy (83.2%) and false-positive rate (FPR 19.0%) for a setup with only six electrodes over the motor cortex and frontal lobe and for an extended 4-38 Hz EEG frequency range while the subject is being stimulated via a median nerve. Configurations with a larger number of electrodes result in higher accuracy (94.5%) and FPR (6.1%) for 128 electrodes (and respectively 88.0% and 12.9% for 13 electrodes). CONCLUSION: The present work demonstrates that using an extended EEG frequency band and a modified EEGNet deep neural network increases the accuracy of MI detection when used with as few as 6 electrodes which include frontal channels. SIGNIFICANCE: The proposed method contributes to the development of Brain-Computer Interface systems based on MI detection from EEG.

Deep Learning Techniques to Improve Intraoperative Awareness Detection from Electroencephalographic Signals.

Every year, millions of patients regain conscious- ness during surgery and can potentially suffer from post-traumatic disorders. We recently showed that the detection of motor activity during a median nerve stimulation from electroencephalographic (EEG) signals could be used to alert the medical staff that a patient is waking up and trying to move under general anesthesia [1], [2]. In this work, we measure the accuracy and false positive rate in detecting motor imagery of several deep learning models (EEGNet, deep convolutional network and shallow convolutional network) directly trained on filtered EEG data. We compare them with efficient non-deep approaches, namely, a linear discriminant analysis based on common spatial patterns, the minimum distance to Riemannian mean algorithm applied to covariance matrices, a logistic regression based on a tangent space projection of covariance matrices (TS+LR). The EEGNet improves significantly the classification performance comparing to other classifiers (p- value <; 0.01); moreover it outperforms the best non-deep classifier (TS+LR) for 7.2% of accuracy. This approach promises to improve intraoperative awareness detection during general anesthesia.