EEG pattern differences in motor imagery based control tasks used for brain-computer interfacing: From training sessions to online control Book in Scopus uri icon

abstract

  • © 2018 Nova Science Publishers, Inc.Brain-computer interfaces (BCIs) are promising systems that attempt to replace the function of the brain output pathways by using the brain signals to control a device of interest. Investigating the control tasks (specifically motor imagery (MI) tasks) used to operate a BCI system under different demanding conditions may explain the difficulty to employ this type of system outside the laboratory. Therefore, the present study set out with the aim of quantifying and qualifying the electroencephalographic (EEG) patterns of commonly used control tasks in BCI systems under different task states. The analysed control tasks were three: left hand MI, right hand MI, and a relaxed but focused mental state. The different task states referred to eight different scenarios, whereby a random sample of eleven participants were guided from modulating their brain signals using MI related control tasks, to use those control tasks for selecting activities of daily living in simulated living situations. The EEG patterns were analysed in line with the EEG features that best differentiated among the three control tasks, and the electrophysiological origin (recording sites, frequency bands, and time windows) of those features. Taken together, the findings of this study highlight the impact of the human brain processing on the BCI system performance. It has been demonstrated that the EEG patterns of MI related control tasks are not only determined by MI activity per se, but they are also defined by the processing of internal (e.g., navigation strategy and decision making) and external (e.g., sensory stimuli or number of tasks to be attended) events associated with the working environment. The investigation of the environmental effects on the user control tasks is very important in order to achieve the desirable overt and covert adaption in BCI systems.

publication date

  • January 1, 2018