Abstract
The development of new approaches to detect motor-related brain activity is key in many aspects of science, especially in brain–computer interface applications. Even though some well-known features of motor-related electroencephalograms have been revealed using traditionally applied methods, they still lack a robust classification of motor-related patterns. Here, we introduce new features of motor-related brain activity and uncover hidden mechanisms of the underlying neuronal dynamics by considering event-related desynchronization (ERD) of μ-rhythm in the sensorimotor cortex, i.e., tracking the decrease of the power spectral density in the corresponding frequency band. We hypothesize that motor-related ERD is associated with the suppression of random fluctuations of μ
-band neuronal activity. This is due to the lowering of the number of active neuronal populations involved in the corresponding oscillation mode. In this case, we expect more regular dynamics and a decrease in complexity of the EEG signal recorded over the sensorimotor cortex. In order to support this, we apply measures of signal complexity by means of recurrence quantification analysis (RQA). In particular, we demonstrate that certain RQA quantifiers are very useful to detect the moment of movement onset and, therefore, are able to classify the laterality of executed movements.
The detection of the motor-related brain activity for noninvasive electroencephalogram (EEG)-based brain–computer interfaces (BCIs) is an actively discussed topic in many areas of research. This is of special interest in the context of neurorehabilitation and non-muscular control of remote devices using BCI-based techniques. Traditionally used methods for motor-related feature extraction, such as spatial filtering and time-frequency analysis, allow one to associate motor actions with event-related desynchronization (ERD) of μ
-band oscillations (8–13 Hz) over the sensorimotor cortex. However, these features, i.e., location of brain activity sources, amplitudes of spectral components, etc., are of strong inter- and intrasubject variability. Moreover, inherent nonstationarity and a poor signal-to-noise ratio of EEG signals strongly complicate the detection and classification of motor-related patterns in single trials. To find new features of the motor-related brain activity, we explore EEG signals from the viewpoint of signal complexity. In particular, we put forward the hypothesis that μ-band ERD causes the reduction of random fluctuations of neuronal activity, resulting in a more regular behavior of EEG signals during motor task accomplishments. With this goal in mind, we apply recurrence quantification analysis (RQA), a nonlinear method that describes the recurrence structure of a system by several quantifiers, in order to examine its complexity and uncover hidden underlying phenomena. Our findings show that certain RQA measures, namely, determinism and recurrence time entropy, allow one to reveal new features associated with neuronal activity complexity reduction. These measures are not only sensitive to the transitions from background to motor-related brain activity but also very useful for distinguishing different types of motor actions (left/right limbs motion), which is valuable in the context of potential BCI applications
