AN IDENTIFICATION METHOD OF PULSE CONDITION OF PATIENTS WITH HYPERTENSION

作者：Wang Yiran^*; Zhang Shaohui; Qi Yingchun

来源：Journal of the Balkan Tribological Association, 2016, 22(2): 1410-1420.

• pathology
• hypertension
• pulse signal
• wavelet transformation modulus
• principle of modulus maxima
• BP neural network

In order to deal with the problem that hypertensive patients are not easy to be found early, the paper proposes a new method, which is based on the wavelet modulus of the maximum principle and BP neural network to analyse pulse signal. The new approach first decomposes pulse signals into three layers of wavelet packet and it locates the mutational sites in the light of the maximum principle. Then we can obtain wavelet diastole and systole to construct feature vectors of pulse signal. These vectors are input into classifiers of BP neural network to do experiments. The experimental results show that normal human recognition rate is 93.3%, and hypertension's is 86.7%. The method can effectively detect mutational points, and correctly recognize pulse signal of normal people and hypertensive patients. Compared with other methods, the paper is the first which combines wavelet modulus and BP neural network to measure blood pressure. It has simple, safe and highly reliable characteristics. This paper proposes the pulse signal identification method for the first time, and it has been verified in clinical practice. The method has high innovation.

• 出版日期2016
• 单位周口师范学院