基于运动学和肌电特征的脑瘫爬行运动功能异常分析与评估

Analysis and evaluation of cerebral palsy crawling abnormalities based on kinematics and electromyography characteristics

目的探索基于运动学和肌电特征的脑瘫患儿爬行运动功能异常量化评估方法。方法以儿童手膝爬行作为研究对象。首先,利用动力学传感器和表面肌电检测技术,提取可有效描述爬行运动强度、稳定性、肌肉激活水平以及拮抗肌对协调性等功能状态特征参数;然后,针对多通道数据特征对爬行运动功能进行整体量化评估时存在的数据冗余现象,利用主成分分析(principal component analysis,PCA)算法进行特征融合,提出基于特征距离比的脑瘫患儿爬行运动异常量化评估方法。结果对17名健康儿童和22名脑瘫患儿开展了爬行运动分析,发现PCA特征融合可有效减少多维特征中的冗余信息,基于融合特征距离比的方法可有效描述脑瘫患儿的爬行运动异常。结论本文提出的基于运动学和表面肌电特征距离比的方法,为脑瘫患儿爬行功能异常的量化评估提供了可能。

Objective To explore the quantitative analysis and evaluation method for crawlingabnormalities in children with cerebral palsy （CP） based on kinetic and surface electromyography （ sEMG）characteristics. Methods Children’s hand.knee crawling was studied in this paper. Firstly, the inertial sensorsand sEMG electrodes were used to capture the features that effectively described the intensity, stability, muscleactivation level and coordination of antagonistic muscle pairs during hand.knee crawling. Secondly, aiming tosolve data redundancy problem in the overall quantitative evaluation of crawling motor function from multi.channel data features, the principal component analysis （PCA） algorithm was used for feature fusion, and thequantitative evaluation method of crawling motion abnormality was proposed based on feature distance ratio.Results There were 17 healthy children and 22 children with CP recruited in this study. The experimentalresults demonstrated that PCA.based feature fusion could reduce the redundant information in multi.dimensionalfeatures, and the evaluation method based on feature distance ratio could effectively describe the abnormalcrawling motion of CP children. Conclusions This method based on kinematics and sEMG feature distance ratioprovides a possibility for quantitative evaluation of crawling abnormalities in children with CP.