Neuroengineering in Brain-Computer Interfaces

Abstract

Brain-computer interfaces translate neural activity into commands for external devices, offering restored communication and motor control to people with severe paralysis, yet the engineering gap between laboratory demonstrations and durable clinical systems remains wide. We assessed 218 neuroengineering BCI programmes conducted across laboratories and clinical sites in Switzerland, Sweden, and France between 2014 and 2024, spanning five platform categories: intracortical microelectrode arrays, electrocorticographic grid systems, scalp electroencephalography decoders, functional near-infrared spectroscopy interfaces, and hybrid multimodal BCI systems. A Neuroengineering BCI Translational Index (NBTI) was constructed from five sub-scores -- decoding accuracy, signal longevity, user training burden, implant safety profile, and real-world usability -- with weights from regression against progression to next-phase clinical deployment. NBTI correlated with translational advancement at r = +0.83 and discriminated advancing from stalled programmes with an AUC of 0.880. Intracortical arrays scored highest (mean NBTI 0.812), while fNIRS interfaces trailed at 0.594. Only 33.5 percent of programmes exceeded the 0.75 NBTI threshold. Decoding accuracy carried the largest regression weight (beta = +0.280), followed by signal longevity (beta = +0.228). These results suggest that next-generation BCIs should prioritise chronic electrode stability and adaptive decoding algorithms that maintain performance as neural signals drift over months and years.