A Wireless Neural Signal Acquisition System for Interventional Brain–Computer Interface

Interventional brain-computer interfaces represent a promising approach for neural signal acquisition by placing electrodes within cerebral blood vessels, offering superior signal fidelity compared to surface recordings while avoiding the risks of direct brain tissue penetration. However, traditional neural recording systems require hardware and wired connections, limiting their applicability for chronic implantation in interventional scenarios. This research presents a novel wireless electroencephalography and electromyography signal acquisition system specifically engineered for interventional brain-computer interface applications. The system integrates three critical modules: electromagnetic induction-based wireless power transfer, high-fidelity signal processing via a 24-bit analog-to-digital converter and Bluetooth Low Energy wireless data transmission using the EFR32BG22 system-on-chip. Experimental validation demonstrates successful acquisition of steady-state visual evoked potentials (SSVEPs) with 78.95% classification accuracy using Support Vector Machine algorithms, and simultaneous electroencephalography-electromyogram (EEG-EMG) recording during motor tasks achieving 73.3% multi-class classification performance. Frequency domain analysis reveals less than 5% deviation compared to g.tec systems across the 0-50 Hz operational bandwidth. In-vivo validation using a sheep model with medical-grade silicone encapsulation successfully captured neural signals from transverse sinus and superior sagittal sinus electrodes during behavioral motor tasks. The biocompatible and miniaturized design enables chronic implantation while maintaining clinical-grade signal quality and real-time communication capabilities, establishing a novel device for interventional brain-computer interface applications.