One-Class Riemannian EEG Classifier to Detect Anesthesia
Résumé
Accidental awareness during general anesthesia (AAGA) is a severe complication of anesthetic care. To prevent AAGAs, electroencephalograms (EEG) began to be employed, but current brain monitors still prove inadequate in detecting AAGAs. We aim to propose a new Brain-Computer Interface (BCI) that monitors the motor cortex via EEG, detects movement intentions (MIs) based on Median Nerve Stimulation (MNS) patterns, and alerts the medical team. Consequently, our first objective is to develop a classifier that distinguishes EEG patterns induced by MNS under two conditions: when a patient is awake vs under GA. Since the latter condition is unavailable presurgery for BCI calibration, we focused on one-class methods. A One-Class Riemannian Minimum
Distance to the Mean trained with the awake data correctly differentiates between these two conditions (test balanced accuracy of 85.44%), significantly better than when the classifier is trained with the beginning of intraoperative data.
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|