Continuous 3D control of a virtual helicopter using a motor imagery based BCI

Abstract

Brain-computer interfaces (BCIs) are devices that allow for thought-based control of computer systems. However, sophisticated control of multi-dimensional BCIs has only recently been achieved in non-invasive systems. The design of these systems has focused on giving users fast, autonomous control that is both intuitive and accurate. Through the use of electroencephalographic recording techniques, sensorimotor rhythms induced from motor imaginations may be captured and a control signal may be characterized. Here we have trained two subjects with an interactive and continuous protocol to modulate their sensorimotor rhythms to control three-dimensions of motion of a virtual helicopter to reach randomly positioned and oriented rings. The subject group acquired 88% of presented targets and achieved as many as 11 consecutive rings in a five-minute period. Subjects learned to fly quickly, continuously and accurately through golden rings positioned and oriented randomly throughout a 3D virtual space.