【作者】 王旭东；

【导师】 王宏；

【作者基本信息】 东北大学 ， 机械工程（专业学位）， 2015， 硕士

【摘要】 现代生物医学和临床应用领域,标准的银/氯化银电极广泛应用于生物电信号的检测。但这种电极使用之前需要对皮肤进行处理,这是一项繁杂的工作并且会造成皮肤过敏等不良反应。为解决这些问题,本文研究了一种基于3D打印技术的用于生物电信号检测的干电极。首先在三维建模软件中设计出电极模型,然后将设计文件载入打印机进行打印。这些电极由ABS塑料和铜线制成,具有机械性能稳定、可重复利用、成本低以及使用方便等特点。在人体电信号检测实验中,这种3D打印干电极表现出良好的性能,足以与银/氯化银电极相媲美。本文在实验室自制生物电信号采集设备的基础上进行了软件算法的开发,在心电信号采集过程中信号不可避免地会受到基线漂移、工频干扰、高频噪声和运动伪迹等干扰,这就需要一种降噪方法在保留心电信号的细节信息的同时去除噪声。本文深入研究了小波变换,分析噪声频带范围,利用小波分解与重构的方法实现了采集信号的噪声去除。为了得到心电信号蕴含的更多信息,本文研究了心电图中的QRS波群检测算法。利用QRS波群具有峰值高、变化快等特点,提出了简便高效的R波波峰检测方法,并将其编写到采集软件程序当中,实现了该系统对心率的实时监测功能。并为心率变异分析打下基础。心率变异率是评估人体生理状态的重要指标,通过它能够估计人体自主神经活动的相关信息,以及交感神经和副交感神经平衡作用情况。本文试图寻找一种方法能够准确地利用心率变异分析实现心律失常的诊断。使用PhysioNet数据库中的数据进行了时域、频域和基本的非线性分析,发现这些方法都不能客观地判定心律失常。在系统地总结了关于心率变异性非线性分析的基础上,本文提出了基于霍夫曼编码计算心电信号平均编码长度的方法。实验证实这种方法能快速而准确地诊断出心律失常。更多还原

【Abstract】 Standard commercial Ag/AgCl electrodes are widely used to monitor biopotential in modern clinic and biomedical applications.But there are inevitable limits:skin preparations before application is cumbersome and irritating.To solve the limitation,in this paper we develop a novel 3D printed dry electrode for biopotential recording.3D capable design software is used to draw the electrode model.Then resulting file is simply loaded on a 3D printer.The electrode is made by acrylonitrile butadiene styrene and copper wire.These electrodes are mechanical stable,reusable,low cost and conveniently applied.Impedance of our electrodes is acceptable.They have good performance in electrocardiograph signal recording compared with wet electrodes and can acquire typical electromyogram signal.Moreover,we develop a software applied to lab-made signal acquisition device,there are inevitably high frequency noise,baseline wander,power frequency interference and motion artifact during measurement,it requires a method to remove the noise while retaining details of ECG signals.By studying wavelet transform and analysing noise frequency domain,wavelet decomposition and reconstruction is proved an effective way.In order to learn more information about ECG,this paper proposes algorithms for QRS cluster detection.These functions,taking advantage of QRS cluster feature,are embedded in acquisition software,so that it would show heart rate in real time.Accurate R wave detection is foundation of HRV(heart rate variability)analysis as well.HRV is an important index to evaluate the physiological state of human body.It is concerned with nerve activity and could reflect balance between sympathetic and parasympathetic nerve.This paper attempts to find an approach for arrhythmia diagnosis accurately using HRV analysis.Measures in time domain,frequency domain and nonlinear domain had been examined with date from PhysioNet database,however,results were not so good as expected.After reviewing the development and applicants of nonlinear dynamics theory,this paper proposed the average coding length of ECG based on Hoffman coding,.A series of experiments showed this method is capable to diagnose arrhythmia with fidelity.更多还原