A 2.5-20kSps in-Pixel Direct Digitization Front-End for ECoG with In-Stimulation Recording.

Aditi Jain, Eric Fogleman, Paul Botros,Ritwik Vatsyayan, Corentin Pochet,Andrew Bourhis, Zhaoyi Liu, Suhas Chethan, Hanh-Phuc Le,Ian Galton, Shadi Dayeh,Drew A. Hall

IEEE Custom Integrated Circuits Conference(2024)

引用 0|浏览2
暂无评分
摘要
Closed-loop neuromodulation promises to enhance treatment for movement disorders, pain, and epilepsy. Advancements in low-im-pedance, high-density recording grids [1] have paved the way for low-noise neural recording systems with high spatial and temporal resolution. However, a conventional high-density neural recording signal path with programmable gain amplifiers (PGAs) and a shared ADC [2] saturates during stimulation because of the high amplifier gain. Due to a fundamental tradeoff with the input high-pass cutoff frequency (for dc electrode offset elimination), it takes hundreds of ms to recover, leading to critical data loss. Recent advances in direct digitization-based analog front-ends (AFEs) overcome this limitation by forgoing the amplifier and directly connecting the electrode to a high dynamic range ADC. Directly using a continuous time $\Delta\Sigma$ mod-ulator (CTDSM) for this application has several notable challenges: slow recovery/instability during artifacts beyond the input range, power and area limitations, and low input impedance $(Z_{\text{in}})$ . We report a 4×2 array of per-pixel 2 nd -order $\Delta\Sigma$ ADCs (including the decimation filter) for ECoG with the fastest (sub-ms) artifact recovery time, ena-bling in-stimulation recording and power-efficient bandwidth scaling.
更多
查看译文
关键词
Electrode,Scalable,Power Efficiency,Neural Signals,High Dynamic Range,Input Range,Neural Recordings,Phase Unwrapping,Varactor,Flicker Noise,Stimulation Artifact,Noise Immunity,Gray Code,Additional Bits,High-density Recordings,Monopolar Stimulation,Commercial Chip,Current Part,Input-referred Noise,Clock Generator
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要