Detection of motor task difficulty level from EEG data

摘要

Rehabilitation protocols are used to increase daily life activities of locked-in patients. There are ongoing efforts to use brain-computer interfaces (BCI) in various ways to increase the benefits of such rehabilitation protocols to patients. An interesting claim is that if a system can detect the intention level of a patient and update the daily program according to this patient's motivation, the gain from these rehabilitation protocol can be increased. In this study, a system that records the electroencephalography (EEG) signals of healthy users performing arm movements against two levels of force has been designed based on the assumption that intention level is proportional to the level of motor task difficulty. EEG signals from 7 healthy subjects and 3 channels were recorded while subjects were performing work against two different levels of force. We calculated frequency bands of these channels and applied linear discriminant analysis (LDA) for classification of two environments corresponding to two motor task difficulty levels and resting state.