基于经验参数和小波变换提取颈动脉波的时域特征

目的:为了解决应用小波变换进行颈动脉波自动检测运算量大的问题,提出一种改进的基于经验参数和小波变换的颈动脉波自动检波算法。方法:首先对脉搏波信号进行小波分解,再将小波分解的某细节信号按基线取绝对值,然后运用小波变换的奇异点检测原理确定前两个有效周期的极大值点位置,接着结合生理知识和实际经验对下一周期的极大值点加以预测,最后回到时域信号中,结合经验参数在一定范围内确定各特征点的位置。结果:经过对比分析发现本文主算法耗用时间比过零点法减少一半以上,较大地提高了运算速度。结论:该方法具有准确、方便、直观、运算量小等优点,由于可以不依赖于心电信号实现脉搏波自定位,因而特别适合单独进行脉搏波分析;结果表明,本方法在保证检测精度的前提下,让运算效率得到较大的改善,利于进行实时分析。

Based on the Experience of Parameters and Wavelet Transform to Extract Carotid Pulse Wave of Time-Domain Characteristics

Objective: To solve the problem of carotid pulse wave automatic detection of the large amount of computation using the wavelet transform, we give an improved carotid pulse wave of automatic detection algorithm based on empirical parameters and wavelet transform. Methods: First, the pulse wave signals are decomposed by wavelet, and we get the absolute value of the wavelet decomposition of a details-signal according to the baseline, then use wavelet transform singular-point detection principles to determine the maximum value of the first two effective cycle, and predict the maximum value point of next cycle with physiological knowledge and practical experience, and finally back to the time domain signal, in a certain range to determine the characteristics of the location with empirical parameters. Results: With comparative analysis, we have found that the time of the main algorithm of our method reduced more than half compare with the time of the method through zero-point, and improved the operation speed largely. Conclusion: The method is accurate, convenient, intuitive and computational advantages of small volume. It is particularly suited for single pulse wave analysis, as it is not dependent on ECG signal from the pulse wave achieve positioning. The results showed that the method in ensuring the accuracy of detection premise, let operational efficiency improved significantly, to conduct real-time analysis.