Denk et al., 2018 hearing-device ERP
为什么重要
路线定位:in-ear and around-ear / hearing-device ERP。
任务或证据:Hearing-device ERP switch
自研用途:Control acoustic/device artifacts in hearable EEG tests.
Evidence Matrix Summary
Field Value Route / hardware Hearing device + right cEEGrid/in-concha Task / evidence base Hearing-device ERP switch Main finding cEEGrid SNR exceeded local in-concha; N100/P300 detected to device setting changes. Key limitation ERP confounded by hardware noise onset. Use for our system Control acoustic/device artifacts in hearable EEG tests.
PDF Download
下载 PDF
静态路径:/papers/15-denk-2018.pdf
Detailed Reading Card
基本信息
年份/出处: 2018, Trends in Hearing, 22.
DOI: 10.1177/2331216518788219.
路线: right-ear hearing-device integration with cEEGrid + in-concha electrodes。
本地文件: library/pdfs_by_category/04_sleep_hearing_real_world/15_2018_denk_et_al_event_related_potentials_measured_from_in_and_around_the_ear_electrodes_integrat.pdf
研究问题
目标是首次评估 ear-EEG 与实时 electroacoustic hearing device 集成后,是否能从 N100/P300 ERP 中识别 hearing-device processing setting switch(Abstract; Intro; PDF pp. 1-2)。
第二目标是比较 around-the-ear cEEGrid 与 in-concha electrodes 的信号质量(Intro/Methods; PDF p. 2)。
硬件系统
听力设备: individualized soft silicone in-the-ear earmold,内含 pickup microphone、loudspeaker 和指向鼓膜的 microphone;Master Hearing Aid platform 实时处理,RME Multiface II,input-output delay 7.8 ms(Acoustic Setup; PDF p. 2)。
EEG: 右耳 10 个 cEEGrid around-the-ear electrodes + 3 个 concha electrodes;Ag/AgCl miniaturized ton electrodes 插入 earmold bores(EEG Setup; Fig. 1; PDF p. 3)。
放大器: SMARTING mobile 24-channel DC EEG amplifier,24-bit,500 Hz;R4a/R4b 为 ground/reference(EEG Setup; PDF p. 3)。
电极点位 / 布局
只分析右耳,即 hearing device 所在侧,以模拟 fully integrated system(EEG Setup; PDF p. 3)。
主要 ERP 通道为 vertical bipolar cEEGrid: ((R2 + R3)/2) - ((R6 + R7)/2)(EEG Analysis; PDF p. 5)。
电极配置比较包括 Grid、Concha、Grid/Concha 三类 reference scheme(Influence of Electrode Position; PDF p. 6)。
实验设计
被试: 17 名正常听力被试,age 28.4 +/- 5.4,10 male、7 female;3 人因行为标准未达标排除,统计分析保留 14 人(Participants; Results; PDF pp. 2, 6)。
任务: 每 trial 四个声音,半数 trial 在第三和第四声之间切换 hearing-device output filter;被试判断第三/第四声是否有差异(Paradigm; PDF p. 4)。
刺激: Noise、Speech、Speech-in-Noise,500 ms,间隔 300 ms;总 192 trials,4 blocks,session 约 90 min(Stimuli/Paradigm; PDF pp. 4-5)。
信号处理流程
EEGLAB + MATLAB,统计用 RStudio;0.1-20 Hz FIR;whole-trial epochs -1000 to 4000 ms,相对单声 epochs -500 to 1000 ms(EEG Analysis; PDF p. 5)。
额外提取 hearing-device hardware noise onset O1-O4 locked epochs;artifact rejection 使用 EEGLAB probability criterion SD=2,平均剔除 27.1% +/- 6.7% trials(EEG Analysis; PDF p. 5)。
N100: peak latency +/-20 ms;P300: 230-430 ms mean amplitude;SNR 用 plus-minus procedure 估计 noise(EEG Analysis; PDF pp. 5-6)。
结果
行为: identical 平均正确 90.0%,deviant 正确 93.3%,3.6% trials 无回应,说明 filter switch 明显可听(Psychophysical Results; PDF p. 6)。
N100 repetition: S1 > S2,F(1,13)=13.22, p=.003;O4 device onset 的 filter-switch N100 显著,F(1,13)=15.36, p=.002(EEG Results; PDF p. 7)。
P300: O4 device onset deviant > identical,F(1,13)=12.25, p=.004(Fig. 7; PDF p. 8)。
电极比较: cEEGrid Grid SNR 最高,M=16.18, SD=6.16,显著高于 local in-concha Concha M=7.55, SD=6.70, t(13)=4.10, p=.001(Fig. 8; PDF pp. 8-9)。
in-concha channels referenced to R4b 彼此高度相关 r=.99;cEEGrid 与 in-concha 相关低,median r=.12,说明信息部分非冗余(Fig. 9; PDF p. 9)。
局限
核心 ERP 差异主要锁定到 device hardware noise onset O4,而非实际声音 S4;说明被试可能基于硬件噪声完成判断(Discussion; PDF pp. 10-11)。
around-ear 与 in-concha 使用不同电极技术,SNR 比较可能受硬件差异影响;作者建议未来用相同电极技术与更低输入噪声放大器(Discussion; PDF p. 12)。
正常听力被试听到硬件噪声,结果不一定直接外推到听损用户(Discussion; PDF p. 11)。
对自研的启发
助听器-EEG 集成实验必须控制 device noise、filter switching artifacts 和声学 cue,否则 EEG 可能揭示的是设备噪声感知。
around-ear electrodes 在该任务下优于 in-concha,但 in-concha 可能提供非冗余信息;混合布局值得保留。
实时听力设备验证应同时报告声学链路延迟、耳模结构、LSL trigger 同步和硬件噪声。
Field Value ID p15_denk_2018_hearing_device_erpsTitle Event-Related Potentials Measured From In and Around the Ear Electrodes Integrated in a Live Hearing Device for Monitoring Sound Perception Year 2018 Category 04_sleep_hearing_real_worldRoute in-ear and around-ear Stage hearing-device ERP Status processedSource integrity okPages 14 OCR status not_needed
Evidence Groups
Group Hits Pages hardware 12 p. 1, p. 2, p. 3 electrode_layout 12 p. 1, p. 2, p. 3 experiment 12 p. 1, p. 2, p. 3 signal_processing 12 p. 1, p. 2, p. 3, p. 4, p. 5 results 12 p. 1, p. 2, p. 4, p. 5, p. 6 limitations 12 p. 2, p. 3, p. 4, p. 5, p. 6, p. 7, p. 8, p. 9
Local Evidence Sources
Source PDF path: US-pdf/denk-et-al-2018-event-related-potentials-measured-from-in-and-around-the-ear-electrodes-integrated-in-a-live-hearing.pdf
Public PDF path: /papers/15-denk-2018.pdf
Categorized PDF path: library/pdfs_by_category/04_sleep_hearing_real_world/15_2018_denk_et_al_event_related_potentials_measured_from_in_and_around_the_ear_electrodes_integrat.pdf
Extracted text path: library/texts/04_sleep_hearing_real_world/15_2018_denk_et_al_event_related_potentials_measured_from_in_and_around_the_ear_electrodes_integrat.txt
Detailed card source: library/DETAILED_PAPER_CARDS_BATCH_3.md
Page-level evidence index: library/EVIDENCE_INDEX.md
Use policy
The public PDF path is served by the Vocs/Vite static asset pipeline. The source, categorized, and text paths remain local evidence references for close reading.
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.