An advanced TMR sensor-based magnetrode for in vivo LFP magnetic field recording

The detection and interpretation of brain signals are crucial for advancing brain-computer interface (BCI) technologies. Local field potential (LFP) signals, reflecting synchronized neuronal ensemble activity, offer insights into coordinated neural function. In this study, we present a miniaturized tunneling magnetoresistance (TMR)-based neural magnetrode optimized for in vivo LFP magnetic recording. The magnetrode achieves a magnetoresistance ratio (145%) and low-field sensitivity (16.59 %/mT), while maintaining low detection limits of 4.8 nT/√Hz at 1 Hz and 140 pT/√Hz at 1 kHz. Noise analysis revealed that reducing bias current and applying high-frequency AC excitation significantly suppress low-frequency 1/f noise. In vitro simulations validate LFP reconstruction capability, and in vivo experiments demonstrate a strong correlation (r = 0.857 ± 0.031, p < 0.01) between magnetic and electrical LFPs. Furthermore, in vitro durability tests in artificial cerebrospinal fluid demonstrated exceptional stability, with negligible signal drift (<0.4% variation in TMR ratio) over a 7-day period. This work establishes the TMR-based magnetrode emerges as a new potential tool for neural interface technologies, with implications real-time BCI and neuropathology research.