Online 3D Motion Decoder BCI for Embodied Virtual Reality Upper Limb Control: A Pilot Study

Abstract

Non-invasive electroencephalogram (EEG) based brain-computer interface (BCI) users aim to achieve three-dimensional (3D) control using only brain signals. Motion trajectory prediction (MTP) is a method that may be used for translating imagined 3D movement into virtual limb control. This process requires the capture of actual kinematic of limb motion trajectory in an experimental setup to perform MTP. Virtual reality (VR) allows for natural, embodied virtual limb feedback and has the potential to create improved experimental BCI training paradigms through an increased presence in applications such as reach target tasks. Here, results are presented from a novel experimental setup and pilot study involving two subjects attempting to control 3D movement of virtual limbs using imagined 3D movement and show that overall, both subjects were able to achieve some level of control with one session achieving a correlation r=0.39 ±0.131, p<0.28. Average accuracy from all sessions was r=0.203 ±0.092, p<0.12, lower than the target threshold of r=0.3. However, both participants reported positively in how present they felt within the experimental setup indicating that there is scope to investigate how this can be adapted to impact training and motivation while using a 3D BCI.