Spinal cord-computer interfaces for neurorehabilitation: Technical challenges and clinical perspectives

Abstract Objective: This Perspective aims to clarify the clinical and engineering scope of spinal cord-computer interfaces (SCCIs) in neurorehabilitation, distinguishing them from adjacent brain-computer interface, neuromodulation, and assistive-technology approaches while identifying the key barriers to translation. Approach: We synthesize evidence from epidural electrical stimulation, spinal bioelectronics, neural decoding, actuator control, brain-spine interfaces, and activity-dependent rehabilitation to define SCCIs as closed-loop or potentially closed-loop spinal interfaces. We then organize the major translational barriers into five interdependent domains: implantation strategy, long-term interface stability, spinal signal integration, actuator coupling, and activity-dependent circuit reconstruction. Main results: Our synthesis identifies five barriers that currently limit SCCI translation: surgical targeting and localization, chronic interface degradation, heterogeneous spinal pathophysiology, real-time decoding and actuator compatibility, and incomplete understanding of activity-dependent circuit remodeling. Although targeted spinal stimulation and hybrid brain-spine systems have shown functional benefits in selected populations, these advances remain difficult to generalize without adaptive, patient-specific system architectures. Significance: By distinguishing SCCIs from adjacent neurotechnology paradigms and linking engineering barriers to clinical decision points, this framework may support consistent terminology, calibrated translational claims, and the development of adaptive, patient-specific closed-loop strategies for restoring native motor control.