Skip to content
cEEGrid Guide

Van Den Broucke et al., 2023

wireless hardware

为什么重要

  • 路线定位:wireless around-the-ear cEEGrid / wireless hardware。
  • 任务或证据:ABR-oriented hardware vs BioSemi
  • 自研用途:ABR needs different front end than OpenBCI/Smarting.

Evidence Matrix Summary

FieldValue
Route / hardwareCustom wireless 16 kHz cEEGrid hardware
Task / evidence baseABR-oriented hardware vs BioSemi
Main finding16 kHz, low-noise wearable design can record ABR-like waveforms from cEEGrid.
Key limitationSingle-subject evaluation; incomplete channel map.
Use for our systemABR needs different front end than OpenBCI/Smarting.

PDF Download

下载 PDF

静态路径:/papers/21-van-den-broucke-2023.pdf

Detailed Reading Card

基本信息

  • 年份/出处: 2023, IEEE EMBC.
  • DOI: 10.1109/EMBC40787.2023.10340180.
  • 路线: wireless/wearable high-sampling auditory EEG hardware + around-the-ear cEEGrid。
  • 本地文件: library/pdfs_by_category/05_open_hardware_benchmarking/21_2023_van_den_broucke_et_al_wireless_and_wearable_auditory_eeg_acquisition_hardware_using_around_the_ear_cee.pdf

研究问题

  • 目标是为 auditory EEG,尤其 auditory brainstem response (ABR),设计可穿戴、无线、低功耗且高采样率的 cEEGrid acquisition hardware(Abstract/Intro; PDF p. 1)。
  • 关键需求是 ABR peak latency 在 1-7 ms,需要至少约 6 kHz 采样率;常见 wearable EEG <= 1 kHz 不足以满足该任务(Intro; PDF p. 1)。

硬件系统

  • 系统包含四个物理组件: A cEEGrid, B preamplifier, C motherboard, D ESP32 microcontroller;设计使用 KiCad,通信/存储协议用 C++/Python(System design; Fig. 1; PDF pp. 1-2)。
  • 电极: around-the-ear cEEGrid,系统采集 8 biopotential channels;作者认为 ABR 场景不一定需要 32/64 cap channels(System design; PDF p. 1)。
  • Preamplifier: 针对 EEG 10-100 uV 与 subcortical auditory response 约 0.2 uV,使用 gain factor 100;每通道 INA828 low-noise op-amps,7 nV/sqrt(Hz),CMRR -110 dB,offset 0.5 uV;包含 driven-right-leg circuit(System design; PDF p. 2)。
  • Motherboard: 双 6000 mAh LiPo batteries,regulators,anti-alias low-pass filters,ADS1299 8-channel 24-bit simultaneous-sampling delta-sigma ADC,sample rate 16 kHz,gain 1,pseudo-differential input,trigger circuit 和 external oscillator(System design; PDF p. 2)。
  • MCU: ESP32-WROOM-32,用 double buffer 从 ADS1299 通过 SPI 接收数据,并以 1-bit SD protocol 写入 SD card;每个 sample 27 bytes,包括 3 status bytes 和 8 channels x 24-bit data(System design; PDF pp. 2-3)。
  • 续航: 两块 6000 mAh lithium-polymer batteries 支持约 3 h continuous measurement(System design; PDF p. 2)。

电极点位 / 布局

  • 验证实验中 cEEGrid 贴在右耳周围,使用双面胶和少量 conductive gel(Evaluation; PDF p. 3)。
  • 论文报告 8-channel around-ear acquisition;Fig. 2 caption 中示例 ABR 曲线为 fourth cEEGrid channel referenced to eighth channel,但没有给出完整 cEEGrid channel naming map;具体 DRL/reference/ground 物理点位未完整展开(System design/Evaluation; PDF pp. 1-3)。

实验设计

  • 参考系统: BioSemi ActiveTwo Mk2,64-channel,24-bit,active electrodes,16384 Hz(Evaluation; PDF p. 3)。
  • 环境: 单名 test subject 在 double-walled electrically shielded listening booth 中坐于 reclining chair(Evaluation; PDF p. 3)。
  • 刺激: 3000 clicks,alternating polarity,monaural presentation;clicks 80 us,100 dB peSPL,平均约 10 Hz,每次静默 jitter 最大 10%;MATLAB 48 kHz 生成 clicks 和 triggers(Evaluation; PDF p. 3)。
  • 目标: 比较 proposed system 与 BioSemi reference 在 ABR wave-I/wave-V、noise floor 和硬件指标上的表现(Evaluation/Results; PDF pp. 3-4)。

信号处理流程

  • 数据从 SD card offline 读取(Evaluation; PDF p. 3)。
  • 滤波: 1500 Hz low-pass 与 100 Hz high-pass,均为 4th-order Butterworth(Evaluation; PDF p. 3)。
  • Epoch: trigger 后前 20 ms;剔除 amplitude 最大的 15% epochs 后平均;用 bootstrap 估计 confidence interval / noise floor(Evaluation; PDF p. 3)。

结果

  • 规格对比: proposed system max sample rate 16000 Hz,BioSemi 16384 Hz;sample rate accuracy 两者均 < 200 ps;sampling skew 25 ps vs < 10 ps;bandwidth -3 dB 为 3400 Hz vs 3200 Hz;input-referred noise 0.5 uVrms vs 2.0 uVrms;CMRR at 50 Hz > 110 dB vs > 90 dB(Table I; PDF p. 3)。
  • 带宽: theoretical bandwidth 3297 Hz,实际 sweep measurement 约 3400 Hz(Results; PDF p. 3)。
  • 噪声: theoretical total noise 0.537 uV,shorted-input measurement 0.503 uV,与设计目标一致(Results; PDF p. 3)。
  • ABR: proposed system 可见 wave-I 和 wave-V,latency 符合 100 dB peSPL click 预期;amplitude 约 0.20 uV 量级(Fig. 2; Results; PDF p. 4)。
  • Noise floor: proposed system + cEEGrid 约 0.08 uV,BioSemi cap reference 约 0.06 uV;BioSemi + cEEGrid 约 0.10 uV,说明差异主要来自 electrode modality 而非 proposed electronics(Results; PDF p. 4)。

局限

  • 主要验证在 electrically shielded booth 中完成,不等于真实居家/户外电磁环境验证(Discussion; PDF p. 4)。
  • 作者提到室外初步 pilot 显示连接稳定且信号在预期范围,但尚未形成完整 outside-lab validation(Discussion; PDF p. 4)。
  • 当前系统仍由四个组件组成,后续需小型化合并;也需加入 real-time feature extraction/analysis,并可通过 daisy-chain ADS1299 扩展更多通道或双耳 cEEGrid(Discussion; PDF p. 4)。

对自研的启发

  • 如果目标包括 ABR 或 subcortical auditory EEG,普通 250-1000 Hz wearable EEG 不够,应把采样率、anti-alias bandwidth、trigger precision 和 storage throughput 作为核心指标。
  • cEEGrid 不只可用于 cortical AAD/ERP,也可作为便携 ABR acquisition interface,但需要高增益、低噪声前端和严格触发同步。
  • 后续产品化方向应优先解决外界电磁干扰、运动伪迹、四组件集成和双耳同步。

Metadata

FieldValue
IDp21_van_den_broucke_2023_wireless
TitleWireless and wearable auditory EEG acquisition hardware using around-the-ear cEEGrid electrodes
Year2023
Category05_open_hardware_benchmarking
Routewireless around-the-ear cEEGrid
Stagewireless hardware
Statusprocessed
Source integrityok
Pages4
OCR statusnot_needed

Evidence Groups

GroupHitsPages
hardware12p. 1, p. 2
electrode_layout7p. 1, p. 2, p. 3, p. 4
experiment2p. 1, p. 4
signal_processing12p. 1, p. 2, p. 3, p. 4
results11p. 1, p. 2, p. 3, p. 4
limitations10p. 1, p. 2, p. 3, p. 4

Local Evidence Sources

  • Source PDF path: US-pdf/Wireless_and_Wearable_Auditory_EEG_Acquisition_Hardware_Using_Around-The-Ear_cEEGrid_Electrodes.pdf
  • Public PDF path: /papers/21-van-den-broucke-2023.pdf
  • Categorized PDF path: library/pdfs_by_category/05_open_hardware_benchmarking/21_2023_van_den_broucke_et_al_wireless_and_wearable_auditory_eeg_acquisition_hardware_using_around_the_ear_cee.pdf
  • Extracted text path: library/texts/05_open_hardware_benchmarking/21_2023_van_den_broucke_et_al_wireless_and_wearable_auditory_eeg_acquisition_hardware_using_around_the_ear_cee.txt
  • Detailed card source: library/DETAILED_PAPER_CARDS_BATCH_4.md
  • Page-level evidence index: library/EVIDENCE_INDEX.md

Close Reading Checklist

  • Verify exact figures, tables, page numbers, and statistics against the local PDF before formal citation.
  • Keep missing parameters as Not reported unless the PDF or supplementary material confirms them.
  • Mark any cross-paper synthesis as interpretation rather than a single-paper claim.