【摘要】 动脉硬化无创检测对于预防心血管事件具有重要意义。考虑到心电信号、脉搏波信号之间的耦合及与动脉硬化之间的内在关联,心电信号的特征参数包括RR间期、QRS波宽度、T波幅度,脉搏波信号特征参数包括峰值数、20%主波宽度、主波斜率、脉率及三个波峰的相对高度,使用主成分分析方法对脉搏波特征数据降维后,对样本是否为动脉硬化病例进行评价。将神经网络和模糊逻辑推理有机结合,利用40组临床心电、脉搏波信号,建立基于自适应神经模糊推理系统(ANFIS)的动脉硬化评价模型。实验结果表明,该模型可通过自学习实现动脉硬化的安全、无创检测。更多还原

【Abstract】 Artery stiffness is a main factor causing the various cardiovascular diseases in physiology and pathology.Therefore,the development of the non-invasive detection of arteriosclerosis is significant in preventing cardiovascular problems.In this study,the characterized parameters indicating the vascular stiffness were obtained by analyzing the electrocardiogram(ECG)and pulse wave signals,which can reflect the early change of vascular condition,and can predict the risk of cardiovascular diseases.Considering the coupling of ECG and pulse wave signals,and the association with atherosclerosis,we used the ECG signal characteristic parameters,including RR interval,QRS wave width and T wave amplitude,as well as the pulse wave signal characteristic parameters(the number of peaks,20% main wave width,the main wave slope,pulse rate and the relative height of the three peaks),to evaluate the samples.We then built an assessment model of arteriosclerosis based on Adaptive Network-based Fuzzy Interference System(ANFIS)using the obtained forty sets samples data of ECG and pulse wave signals.The results showed that the model could noninvasively assess the arteriosclerosis by self-learning diagnosis based on expert experience,and the detection method could be further developed to a potential technique for evaluating the risk of cardiovascular diseases.The technique will facilitate the reduction of the morbidity and mortality of the cardiovascular diseases with the effective and prompt medical intervention.更多还原