Engineering Neural Interface Materials for Brain-Computer Interfaces: Toward Reliable Neural Information and Functional Recovery

Motor dysfunction is a major disabling outcome of neurological disorders, including stroke, spinal cord injury, and cerebral palsy. By establishing direct communication between neural activity and external effectors, brain-computer interfaces (BCIs) offer a promising strategy for neural functional reconstruction, enabling motor intention decoding, assistive control, and rehabilitation feedback. However, sustained BCI performance depends not only on advances in decoding algorithms but also on neural interface materials and interfacial properties, as well as the reliable transfer of neural information across signal acquisition, intelligent decoding, control execution, and closed-loop feedback. Existing reviews have mainly focused on invasiveness, electrode technologies, or decoding methods, with comparatively limited integration of neural interface materials, signal reliability, artificial intelligence (AI) adaptability, and closed-loop performance within a unified information reliability framework. Therefore, this review summarizes recent advances in AI-enabled BCIs for motor rehabilitation, compares neural signal and interface characteristics across different levels of invasiveness, examines the effects of key material and interfacial properties on stable signal acquisition, and discusses the roles of AI in data-efficient learning, spatiotemporal and relational neural representation, adaptation to neural distribution shifts, and efficient decoding for real-time closed-loop operation. Representative applications in exoskeletons, neuroprostheses, cerebral palsy rehabilitation, and rehabilitation robotics are further summarized. Finally, current challenges and potential solutions spanning neural interface stability to functional recovery are discussed, providing a systematic pathway toward long-term reliable BCIs and the clinical translation of next-generation intelligent neurorehabilitation systems.