A Novel Brain-Computer Interface Approach Designed for Dynamic Target Selection

This paper presents a dynamic brain-computer interface (BCI) approach for multiple mobile targets selection in complex scene. In this study, two dynamic paradigms based on the P300 and the steady-state visually evoked potential (SSVEP) respectively were devised and tested by experiments. In the experiments, a video in which several people walks around in a laboratory is presented on the screen, with the random flashes or periodic flickers at different frequencies superimposed on the mobile objects. Then subjects were told to select the target using these paradigms. Eight subjects participated in the experiments. During the P300 session, each subject performed 48 online P300 tasks. They achieved an average accuracy of 85.67%. In the SSVEP session, the optimal stimulation time for each subject was selected through different sessions, and the average accuracy and response time were 83.6% and 3.875s, respectively. The study indicated the feasibility and efficiency of dynamic target selection BCI approaches, and the BCIs we designed can be used in other fields to extend the scope of applications.