融合生物运动信息的上肢运动功能康复诊断方法研究  

Research of Diagnostic Method for Upper Limb Motor-Function Based on Fusion of Biological Motion Information

作　　者：高云园[1] 高发荣[1] 罗志增[1] 

机构地区：[1]杭州电子科技大学自动化学院智能控制与机器人研究所,杭州310018

出　　处：《中国生物医学工程学报》2013年第6期692-698,共7页Chinese Journal of Biomedical Engineering

基　　金：国家自然科学基金(61201300);国家自然科学基金(61201302);浙江省自然科学基金(LY12F03006)

摘　　要：针对上肢运动功能的康复诊断问题,提出一种融合运动力学信息与生物电信息的综合性上肢运动功能评价方法,用于评价运动功能障碍患者肩肘腕关节的运动功能水平。在上肢进行动作时,对加速度信号和肌电信号进行信号采集、有效信号的选择、特征提取和特征筛选,并利用两类信号在动作分析中的不同优势,将信号的特征值进行优化组合。以简式Fugl-Meyer评分值为标准,构建多个线性回归模型,实现上肢运动功能的综合性评价。在对10位受试者肩肘腕的7个上肢动作(握拳、展拳、屈腕、伸腕、屈肘、伸肘、上肢平举)功能诊断实验中,提出的诊断方法不仅可以进行实时的信息提取和功能诊断,而且与Fugl-Meyer评价有很强的一致性,相关系数达99%以上。实验表明,该诊断方法能取代传统的上肢运动功能评价方法,并更细致地对上肢运动功能进行量化评分。A comprehensive method, fused with kinematics information and bioelectricity information, for upper limb motor-function evaluation was proposed. The method was aimed to evaluate the motor functions of the shoulder, elbow and wrist joints of the hemiplegic patients objectively and quantitatively. Acceleration signal and EMG signal were collected and selected, and then feature extraction and feature selection were done during the movement of the upper limb. Utilizing the distinct superiorities of the two signals, characteristic value of the signals was optimally combined. Several linear regression models were constructed to realize comprehensive assessment of the upper limb motor-function based on Fugl-Meyer assessment. In the function diagnostic experiment of 7 movements of the shoulder, elbow and wrist （namely hand grasps, hand extension, wrist flexion, wrist extension, elbow flexion, elbow extension, front raise） with 10 participants, the evaluation method not only could realize information extraction and function diagnosis in real time, but also had a strong consistency with Fugl-Meyer assessment with a correlation coefficient above 99%. The above experimental results showed that the diagnostic method could quantify the upper limb motor-function more detailed instead of traditional assessment method.

关 键 词：上肢运动功能评价 运动力学信息 肌电信号 特征提取 康复 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]