摘要
不同等级偏瘫患者的表面肌电信号(sEMG)受噪声影响不同,研究适合从偏瘫患者的肌电信号中检测肌肉活动的算法.对Brunnstrom分级Ⅰ-Ⅴ级偏瘫患者,采集双侧共同腕伸运动时前臂原动肌的肌电信号,将健康侧的信号作为对照组.采用运动/静息比方法,计算信号信噪比(SNR),对信号进行绝对值均值(MAV)、模糊熵(FuzzyEn)、样本熵(SampEn)、近似熵(ApEn)的滑动窗运算,比较在不同等级患者中各特征算法的优劣.在不同等级偏瘫患者中,患侧肌电信号的SNR与患者等级呈正相关性.与MAV法相比,3种熵值算法对Ⅱ-Ⅴ级偏瘫患者sEMG运动检测的适应性更好,有检测弱肌力患者潜在运动信号的潜力,其中FuzzyEn比其他熵值算法的适应性更好.对噪声的敏感性方面,FuzzyEn受影响最小.
A suitable algorithm was analyzed to detect muscle activity from surface electromyography(sEMG)feature of hemiplegic patients aiming at the problem that the sEMG signals of hemiplegic patients with different stages were differently affected by noise.EMG signals were recorded from bilateral forearm agonists of BrunnstromⅠ-Ⅴstage patients,while they were implementing wrist extension.The healthy side signals were served as control groups.The signal-to-noise ratio(SNR)of sEMG was calculated based on the ratio of motion to resting signals.Then mean absolute value(MAV),fuzzy entropy(FuzzyEn),sample entropy(SampEn)and approximate entropy(ApEn)were calculated in a sliding window of signals.The characteristics of these algorithms were analyzed in different stage patients.The SNR of sEMG signals in the affected limb of hemiplegic patients is positively correlated with the patients stage.Three entropy algorithms are more suitable in detecting weak sEMG signals from hemiplegic patients of stageⅡ-Ⅴthan MAV algorithm,and have the ability to detect intentional muscle activities.FuzzyEn is better than the other two entropy algorithms.Since the four features have different sensitivities to spike noise or background noise,FuzzyEn is less affected than SampEn,ApEn,MAV.
作者
赵翠莲
徐浩宇
罗林辉
王凯
ZHAO Cui-lian;XU Hao-yu;LUO Lin-hui;WANG Kai(School of Mechatronic Engineering and Automation,Shanghai University,Shanghai 200072,China;Shanghai Jing-an Geriatric Hospital,Shanghai 200040,China)
出处
《浙江大学学报（工学版）》
EI
CAS
CSCD
北大核心
2018年第4期798-805,818,共9页
Journal of Zhejiang University：Engineering Science
基金
上海市科学技术委员会资助项目(16441909000
10441900802)
作者简介
赵翠莲(1963—),女,教授,从事生机电一体化与康复工程研究.orcid.org/0000-0001-9957-7958.E-mail:clzhao@shu.edu.cn
