Frequency optimization for electrodes in implantable brain-computer interfaces

Abstract

Fully implanted brain-computer interfaces (BCIs) are preferred as they eliminate signal degradation caused by interference and absorption in external tissues, a common issue in non-fully implanted systems. To optimize the design of electroencephalography electrodes in fully implanted BCI systems, this study investigates the penetration and absorption characteristics of microwave signals in human brain tissue at different frequencies. Electromagnetic simulations are used to analyze the power density distribution and specific absorption rate (SAR) of signals at various frequencies. The results indicate that lower-frequency signals offer advantages in terms of power density and attenuation coefficients. However, SAR-normalized analysis, which considers both power density and electromagnetic radiation hazards, shows that higher-frequency signals perform better at superficial to intermediate depths. Specifically, at a depth of 2 mm beneath the cortex, the power density of a 6.5 GHz signal is 247.83% higher than that of a 0.4 GHz signal. At a depth of 5 mm, the power density of a 3.5 GHz signal exceeds that of a 0.4 GHz signal by 224.16%. The findings suggest that 6.5 GHz is optimal for electrodes at a depth of 2 mm, 3.5 GHz for 5 mm, 2.45 GHz for depths of 15-20 mm, and 1.8 GHz for 25 mm.

Keywords

brain-computer interfaces / electromagnetic simulation / electroencephalography electrodes / power density / specific absorption rate

Funding

The Open Project of State Key Laboratory of Smart Grid Protection and Operation Control in 2022(SGNR0000KJJS2302150)