Restoring Moderate to Severe Loss of Upper Extremity Function Post-stroke with a Neuromotor Prosthesis: A Case Report

Research Objectives To study the effect of a wearable, powered exoskeleton orthosis controlled by a brain-computer interface (BCI) on upper extremity function in a single participant with moderate to severe hemiparesis. Design Single subject design with pretest, mid, and posttest. Setting Home and outpatient. Participants Participant was a 41-year old right-hand-dominant male with moderate to severe (Fugl-meyer upper extremity [FM-UE] 33/66) left-sided hemiparesis due to a stroke that occurred 18 months prior, with left homonymous hemianopsia, left hemi-neglect, left central VII nerve palsy, and spasticity in the weakened left upper extremity. The stroke lesioned the right lentiform nucleus and adjacent white matter. Interventions Brain implants were surgically placed within the right precentral gyrus region of the cortex, representing the left hand. The 12-week intervention consisted of neuromotor prosthesis training with BCI controlling the powered exoskeleton and biofeedback for controlling a game cursor; physical therapy once/week; occupational therapy twice/week. Main Outcome Measures Action Research Arm Test (ARAT), FM-UE, Motor Activity Log (MAL) domains of Amount of Use (AOU) and Quality of Movement (QOM) Results ARAT scores with the powered exoskeleton were greater with BCI (change: 10-point, 18%) than with myoelectric control (change: 5-points, 9%). Upper extremity outcomes exceeded the minimal clinically important difference thresholds for FM-UE (change: 8-point, 21%), MAL-AOU (change: 18.5-points, 12%), and MAL-QOM (Change: 17-point, 11%). Conclusions The results indicate promising outcomes of the neuromotor prosthesis on upper extremity function for a single participant with potential for future research with implanted systems combined with therapy. Author(s) Disclosures Drs. Serruya and Napoli are inventors on a US provisional patent application that has been filed by Thomas Jefferson University on the software methods used in this study to map neural activity into orthosis control.