A Low-Power Wireless 128-Channel Neural Interface Circuit for Multiregional Brain Recording

作者：Yahao Song, Tianyi Liu, Chao Sun, Yuwei Zhang, Minqian Zheng, Milin Zhang · 年份：2024 · DOI：10.1109/biocas61083.2024.10798169 · 被引用次数：1 · 研究领域：Neuroscience and Neural Engineering、EEG and Brain-Computer Interfaces、Advanced Memory and Neural Computing

This paper proposes a low-power, wireless 128channel neural interface circuit, comprising a 40 nm recorder circuit and a WiFi module. The proposed design enables realtime monitoring and neural signal decoding during behavioral studies, facilitating high-density, long-term tracking of neural signals in freely moving subjects. The recorder circuit integrates 16 -bit resolution at a maximum rate of 32 k samples per second (sps), with a power consumption of 3.43 mW and an area of $6.27 \mathrm{~mm}^{2}$,achieving the lowest normalized power consumption of $0.84 \mu \mathrm{~W} / \mathrm{ch} / \mathrm{ksps}$ among state-of-the-art works. The VoltageControlled Oscillator (VCO) Analog Front Ends (AFEs) feature an input-referred noise of $0.66 \mu \mathrm{~V}_{\text {rms}}$ within the $0.5-60 \mathrm{~Hz}$ range. The digital backend allows flexible channel configuration and includes an on-chip CIC filter to reduce out-of-band noise. The WiFi module, connected via QSPI, facilitates wireless control and data transmission. The proposed circuit was validated through electrocorticogram (ECoG) measurements in rodent subjects, demonstrating its reliability and flexibility.