基于神经网络的脉搏波信号血压检测算法

脉搏波信号蕴含大量的人体生理与病理信息，与血压的变化息息相关，利用其特征参数可以无创连续检测血压。神经网络因其极强的学习能力、泛化能力以及可以充分逼近任意复杂的非线性关系而被应用于脉搏波血压提取算法中。本研究介绍了脉搏波特征参数，并简述了基于脉搏波特征参数进行血压测量的研究进展，详细叙述了基于神经网络的脉搏波特征参数血压检测算法，最后对不同神经网络模型的优缺点进行分析，并对基于神经网络的脉搏波特征参数血压监测算法的研究方向进行展望。

Neural network-based blood pressure detection algorithms for pulse wave signals

Abstract: The pulse wave signal contains a large amount of human physiological and pathological information which is closely related to the changes of blood pressure. The blood pressure can be measured continuously and noninvasively using the characteristic parameters of pulse wave signals. Neural network is used in pulse wave blood pressure extraction algorithm because of its strong learning ability, generalization ability and the ability to fully approximate any complex nonlinear relationship. Herein the characteristic parameters of pulse wave are introduced. The advances in research of blood pressure measurement based on the pulse wave characteristic parameters are briefly described, and the neural network-based blood pressure detection algorithm using the pulse wave characteristic parameters is described in details. Finally, the advantages and disadvantages of the network model are analyzed, and the research direction of the neural network-based blood pressure monitoring algorithm using the pulse wave characteristic parameters is discussed.