MRI Compatible Valve Enables Fast Actuation of Soft Hand Exoskeleton in Medical Imaging

Medical imaging offers a comprehensive approach to studying complex sensorimotor processes in the brain, like motor learning and plasticity. While neural correlates of human motor function have been extensively studied using functional Magnetic Resonance Imaging (fMRI), there is a lack of research on neural activity during physical human-machine interaction, e.g., exoskeleton use. A major factor for this research gap is that integrating exoskeletons into MRI experiments raises compat-ibility challenges, including safety concerns and susceptibility artifacts due to metallic and electrical components. To overcome this problem, all system metal components should be placed in the MRI control room. However, this introduces unwanted delay into the system and makes the exoskeleton move slower than natural movements. We present a novel MRI-compatible exoskeleton system, including MRI-compatible valves and an MRI-compatible exoskeleton glove, developed for low-delay sen-sorimotor experiments inside an MRI scanner. Our results show that using the MRI-compatible valves doubles the grasping speed of the exoskeleton when compared to operating classical valves in the MRI control room. We furthermore show that the system does not produce strong imaging artifacts when a healthy participant uses the low-delay exoskeleton system inside the scanner (represented by a high mean voxel SNR of 283,380) and that their brain activity during rhythmic hand-opening-and-closing with the low-delay design was more similar to that of natural grasping.