On the Neural Correlates of Motor Imagery With an Extra Virtual Arm

Abstract

Motor augmentation (MA) is an emerging field at the intersection of engineering, robotics, and neuroscience, aiming to enhance human capabilities through the integration of extra limbs. This concept leverages the body’s physiological redundancies, including those within the nervous system. This study examined motor imagery (MI) involving a virtual extra arm, focusing on differentiating its neural patterns from those of biological limbs. Thirty participants performed unimanual reaching MI tasks before (Pre) and after (Post) a conditioning phase in a virtual environment, during which half of the participants received tactile feedback on the movement of the extra arm. Electroencephalographic (EEG) recordings revealed distinct event-related desynchronization (ERD) in α and β rhythms between the extra and biological limbs. Additionally, a Riemannian decoder successfully classified MI for the left, right, and extra virtual arm, providing further evidence of distinct neural patterns. While the conditioning played a role in the ERD’s neural signatures, we did not find the same effects on the decoding. We believe that more complex movements, other sensory encoding modalities, or longer conditioning periods would likely strengthen the connection between tactile feedback and neural control.