A lack of anesthesia exposes the patient to awakening during surgery; conversely, an excess of anesthesia increases the risk of postoperative cardiac and neurological complications. A precise optimization of the depth of anesthesia must therefore be sought, in particular for fragile, elderly or co-morbid patients. Our system takes into account the data from the usual monitoring of patients and by integrating them at a high frequency during general anesthesia, allows to deduce the depth of anesthesia. In an alternative model, we propose to add the analysis results of an EEG channel to the model to significantly increase the accuracy of the prediction.
Keywords: Machine Learning, Predictive algorithm, Individual medicine, Depth of anesthesia, Multimodal monitoring, General anesthesia