Decoding motor responses from the EEG during altered states of consciousness induced by propofol
Journal of Neural Engineering February 9, 2016 Y. Blokland, J. Farquhar, J. Lerou et al.
A brain-computer interface (BCI) can detect movement attempts from EEG signals with high accuracy even under propofol sedation. At a propofol concentration of 0.5 μg ml⁻¹, mean classification accuracy was 85% (95% CI 81%–89%), and a classifier trained before sedation and tested during sedation achieved 83% (79%–88%). At 1.0 μg ml⁻¹, accuracies were 81% (76%–86%) and 72% (66%–79%), respectively. In four subjects at the highest concentration, movement-related brain responses largely diminished and transfer classification accuracy was not significantly above chance; these subjects showed slower, more erratic task responses, indicating an altered state of consciousness. The findings suggest BCI technology could detect intra-operative awareness, but the relationship between motor responses and consciousness requires further investigation.