Decoding covert speech for intuitive control of brain-computer interfaces based on single-trial EEG: a feasibility study.

For individuals with severe motor deficiencies, controlling external devices such as robotic arms or wheelchairs can be challenging, as many devices require some degree of motor control to be operated, e.g. when controlled using a joystick. A brain-computer interface (BCI) relies only on signals from the brain and may be used as a controller instead of muscles. Motor imagery (MI) has been used in many studies as a control signal for BCIs. However, MI may not be suitable for all control purposes, and several people cannot obtain BCI control with MI. In this study, the aim was to investigate the feasibility of decoding covert speech from single-trial EEG and compare and combine it with MI. In seven healthy subjects, EEG was recorded with twenty-five channels during six different actions: Speaking three words (both covert and overt speech), two arm movements (both motor imagery and execution), and one idle class. Temporal and spectral features were derived from the epochs and classified with a random forest classifier. The average classification accuracy was $67 \pm 9$ % and $75\pm 7$ % for covert and overt speech, respectively; this was 5-10 % lower than the movement classification. The performance of the combined movement-speech decoder was $61 \pm 9$ % and $67\pm 7$ % (covert and overt), but it is possible to have more classes available for control. The possibility of using covert speech for controlling a BCI was outlined; this is a step towards a multimodal BCI system for improved usability.

Feature domain-specific movement intention detection for stroke rehabilitation with brain-computer interfaces.

Brain-computer interface (BCI) driven electrical stimulation has been proposed for neuromodulation for stroke rehabilitation by pairing intentions to move with somatosensory feedback from electrical stimulation. Movement intentions have been detected in several studies using different techniques, with temporal and spectral features being the most common. A few studies have compared temporal and spectral features, but conflicting results have been reported. In this study, the aim was to investigate if complexity measures can be used for movement intention detection and to compare the detection performance based on features extracted from three different domains (time, frequency and complexity) from single-trial EEG. Two data sets were used where four different isometric palmar grasps or dorsiflexions were performed while continuous EEG was recorded. 39 healthy subjects performed or imagined these movements and 11 stroke patients attempted to perform the movements. The EEG was pre-processed and divided into two epoch classes: Background EEG (2 s) and movement intention (2 s). To obtain an estimated detection performance, temporal, spectral and complexity features were extracted and classified (linear discriminant analysis) after the feature vector was reduced using sequential forward selection. The results show that accuracies between 82-87% and 74-80% are obtained for foot and hand movements, respectively. The temporal feature domain was the most dominant for foot movement intention detection, while the spectral features contributed more to the hand movement detection. The complexity features could be used to detect movement intentions, but the performance was much lower compared to temporal and spectral features.