【作者】 周宽久； 陈雪峰；

【Author】 ZHOU Kuan-jiu, CHEN Xue-feng (Institute of Systems Engineering, Dalian University of Technology, Dalian 116024, China)

【机构】 大连理工大学系统工程研究所；

【摘要】 脉搏波的动态特征是评价人体心血管系统生理病理状态的重要依据。研究脉搏波形成的机理和波形变化规律有助发现心脏生理病理变化情况,波形特征提取是研究基础。本文提出采用高斯函数叠加逼近方法对脉搏信号进行时域分析,并利用遗传算法从采样样本中获取叠加函数的特征参数,克服了传统参数估计算法中的局部收敛问题。特征提取研究为下一步的脉象分类研究打下基础,从而能够实现无创地预测分析和诊断心血管疾病。实验结果表明:该算法较为准确地对脉搏波形进行逼近,获取的参数能代表典型脉象的特征。更多还原

【Abstract】 Dynamic characteristics of pulse waveforms are important factors to evaluate cardio-vascular performance of human being. Researches on the formation mechanism of the pulse can help us to find the cardio-vascular laws of diversification. Hence, characteristics of pulse are the foundation of researches. A new algorithm is issued to extract Characteristics of Pulse Waveforms with a combination of three Gauss functions to fit the waveforms and with the genetic algorithm to acquire nine parameters of the three Gauss functions, which are useful to make diagnosis cardiovascular disease without ravage correctly and quickly. Test results show that the algorithm is effective and the nine parameters acquired by the algorithm can exactly express some typical pulse conditions.更多还原