脑电信号中人工耳蜗伪迹去除技术的研究进展  

Research Progress on the Removal of Cochlear Implant Artifacts from EEG Signals

作　　者：倪广健 郑琪[1,4] Ni Guangjian;Zheng Qi(Academy of Medical Engineering and Translational Medicine,Tianjin University,Tianjin 300072,China;Haihe Laboratory of Brain-Computer Interaction and Human-Machine Integration,Tianjin 300392,China;State Key Laboratory of Advanced Medical Materials and Devices,Tianjin University,Tianjin 300072,China;School of Precision Instrument and Opto-Electronics Engineering,Tianjin University,Tianjin 300072,China)

机构地区：[1]天津大学医学工程与转化医学研究院,天津300072 [2]脑机交互与人机共融海河实验室,天津300392 [3]天津大学先进医用材料与医疗器械全国重点实验室,天津300072 [4]天津大学精密仪器与光电子工程学院,天津300072

出　　处：《天津大学学报(自然科学与工程技术版)》2025年第5期441-451,共11页Journal of Tianjin University:Science and Technology

基　　金：国家重点研发计划资助项目(2023YFF1203500);国家自然科学基金资助项目(81971698).

摘　　要：人工耳蜗(CI)作为一种先进的神经假体,为重度至极重度听力损失患者提供了听力恢复的可能性.然而,目前CI患者植入后临床听觉康复评估结果呈现显著差异性.这种差异可能源于植入后大脑功能的重新组织及其多重因素综合作用.因此,研究人员正在努力开发客观评估方法,以准确评估人工耳蜗植入后大脑的可塑性.脑电图(EEG)由于其与CI的良好兼容性、操作的便捷性和非侵入性等特点,已成为一种具有重大临床应用前景的听觉康复客观评估手段.然而,CI的电刺激产生的大量伪迹对EEG信号的质量构成了严重干扰.这些伪迹使神经响应扭曲,难以从EEG记录中提取出准确可靠的大脑活动数据.目前,这一问题已成为利用脑电技术进行临床应用的主要障碍.以往研究中已尝试采用若干技术来降低伪迹的影响,但至今相关方法仍缺乏共识,且这些方法在广泛临床应用中的适应性仍有待验证.本文首先对CI伪迹的特性及其在EEG信号中的表现进行了详细介绍,进而全面回顾了目前脑电图中减少CI伪迹的处理技术进展,包括线性和非线性方法、时频分析技术以及基于统计和机器学习的算法等.进一步探讨了不同信号处理技术在实际应用中的效果和适应性,总结了这些技术的优势与局限性,对技术未来发展提出了展望.Cochlear implants(CIs)are advanced neural prostheses that can restore hearing in patients with severe to profound hearing loss.However,existing research reveales a significant variation in post-implantation clinical auditory rehabilitation assessment results in patients with CIs.This variation may stem from the reorganization of brain function after implantation and the interplay of multiple factors.This has facilitated the need to develop objective assessment methods to accurately evaluate brain plasticity after CI implantation.Electroencephalography(EEG)possesses favorable characteristics,such as compatibility with CIs,ease of operation,and noninvasiveness.Thus,it has emerged as a promising tool for objective assessment in auditory rehabilitation with significant clinical application prospects.However,the abundance of artifacts generated by CI electrical stimulation significantly affects the quality of EEG signals.These artifacts lead to distorted neural responses,which makes the extraction of accurate and reliable brain activity data from EEG records challenging.This has become a major constraint in the clinical application of EEG technology.Although previous studies have attempted to mitigate the impact of these artifacts using various techniques,there is still a lack of consensus on these methods,and their adaptability to widespread clinical applications remains unexplored.This study first introduces the characteristics of CI artifacts and describes their manifestation in EEG signals.Then,it comprehensively reviews the current progress in processing techniques for reducing CI artifacts in EEGs,including linear and nonlinear methods,time-frequency analysis techniques,and algorithms based on statistics and machine learning.In addition,this study analyzes the effectiveness and adaptability of different signal-processing technologies in practical applications,summarizes their advantages and limitations,and provides an outlook on the future development of these technologies.

关 键 词：人工耳蜗 脑电图 伪迹 去噪 听觉客观评估 信号处理 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] R318[自动化与计算机技术—计算机科学与技术]