Potential EEG biomarkers of sedation doses in intensive care patients unveiled by using a machine learning approach

Abstract

Objective. Sedation of neurocritically ill patients is one of the most challenging situation in ICUs. Quantitative knowledge on the sedation effect on brain activity in that complex scenario could help to uncover new markers for sedation assessment. Hence, we aim to evaluate the existence of changes of diverse EEG-derived measures in deeply-sedated (RASS-Richmond agitation-sedation scale −4 and −5) neurocritically ill patients, and also whether sedation doses are related with those eventual changes. Approach. We performed an observational prospective cohort study in the intensive care unit of the Hospital de la Princesa. Twentysix adult patients suffered from traumatic brain injury and subarachnoid hemorrhage were included in the present study. Long-term continuous electroencephalographic (EEG) recordings (2141 h) and hourly annotated information were used to determine the relationship between intravenous sedation infusion doses and network and spectral EEG measures. To do that, two different strategies were followed: assessment of the statistical dependence between both variables using the Spearman correlation rank and by performing an automatic classification method based on a machine learning algorithm. Main results. More than 60% of patients presented a correlation greater than 0.5 in at least one of the calculated EEG measures with the sedation dose. The automatic classification method presented an accuracy of 84.3% in discriminating between different sedation doses. In both cases the nodes' degree was the most relevant measurement. Significance. The results presented here provide evidences of brain activity changes during deep sedation linked to sedation doses. Particularly, the capability of network EEG-derived measures in discriminating between different sedation doses could be the framework for the development of accurate methods for sedation levels assessment.

Data augmentation for deep-learning-based electroencephalography

“…Hence, not only is it well outside the scope of real-time analysis, necessitating offline training, but the statistics of the relevant brain activity may change considerably by the time the model is trained. There have been some attempts at adaptive machine-learning techniques to better track the changing statistics of the signal [9][10][11]. If that is not enough, EEG is generally recorded using tens to hundreds of electrodes recording simultaneously at hundreds or thousands of samples per electrode, whereas a typical dataset, at least in cognitive neuroscience when looking at discrete experimental events, contains only some hundred to a few thousand samples (i.e., experimental trials) at the most.…”

Necessity of Quantitative EEG for Daily Clinical Practice

“…Then, some procedures require relatively simple and involve straightforward methods as FFT [26,34,[64][65][66] and synchronicity measurements [67,68], approaches that clinicians are familiar with. In contrast, other numerical methods are more complex, or the mathematical approach deviates from physiological assumptions [69][70][71][72][73][74], and this probably takes neurophysiologists out of their comfort zone.…”

Utility and rationale for continuous EEG monitoring: a primer for the general intensivist

“…For example, a relatively high rate of unintended burst suppression occurs in critically ill and peri-procedural patients managed with IV anesthetics [ 190 – 193 ]. In a study of 26 critically ill adults, monitored with EEG after TBI and SAH, patients were given deep sedation to a Richmond Agitation-Sedation Scale score of − 4 or − 5; most demonstrated a correlation of sedation dosing with one or more EEG indices [ 194 ]. An automatic classifier had 84.3% accuracy in discriminating between different sedation doses.…”