基于脑血氧信号的下肢运动状态识别方法研究

Research on the Recognition Method of Lower Limbs Movement State Based on Cerebral Blood Oxygen Signal

为了提高下肢运动辅助设备的智能性,准确地识别使用者的运动意识是至关重要的。采用血氧含量信息作为识别下蹲、行走和起立3种下肢运动状态的信号来源,对基于脑血氧信号的下肢运动状态识别方法进行了研究。总共有14位被试对象参与了本次自发运动的实验,运动的开始、运动的结束和运动时长都是由被试自主控制的,在每一个真正的运动任务开始前都会有一段时间的想象运动。实验过程中的信号都是由近脑红外成像设备记录。分析过程中采用小波包变换将频域段分成5个频段:0~0.03 Hz、0.03~0.06 Hz、0.06~0.09 Hz、0.09~0.12 Hz、0.12~0.15 Hz,想象阶段的合氧血红蛋白的均值和变化率被用作分析特征的数据样本。通过单变量特征优化法选择了103个特征向量,采用逻辑回归法、k-近邻法和支持向量机3种子分类器融合的决策模板法对信号进行识别,识别率达到了86.78%。由于所有的特征采用的是想象阶段的数据,时间上具有提前性,所以该结果对于发送指令给辅助运动器械控制患者运动将有非常大的优越性。

In order to improve the intelligence of lower limbs motion auxiliary equipment,it is very important to correctly identify the motion consciousness of users.By using the blood oxygen content information as the signal source to identify the lower limbs motion states of the lower jaw,walking and standing,the identification method of lower limb motion state based on cerebral blood oxygen signal is studied.A total of 14 subjects participate in the experiment of this spontaneous movement,and the beginning,the end and the duration of the exercise are controlled by the subjects themselves,and there is a period of imaginary movement before each real exercise task.The signals during the experiment are recorded by the near-brain infrared imaging device.In the analysis process,the wavelet packet transform is used to divide the frequency domain into five frequency bands:0~0.03 Hz,0.03~0.06 Hz,0.06~0.09 Hz,0.09~0.12 Hz,0.12~0.15 Hz,and the mean and change rate of aerobic hemoglobin in the imaginary stage are used as data samples for analytical features.103 feature vectors are selected by the single-variable feature optimization method,and the decision template combinating the three subs classifiers of logical regression,k-nearest neighbors and support vector machine is used to identify the signals,and the recognition rate researches 86.78%.Since all the features are based on the data of the imaginary phase,which is advanced in time,the results will have a very great advantage for sending instructions to the auxiliary exercise device to control the movement of the patients.