The Study Of The Application Of Neural Networks To The Detection Of Pulse Signals

The diagnostic methods and particularly curative effect of traditional Chinese medicine have been playing an important role of national health care. Pulse-feeling is one of the primary diagnostic methods in traditional Chinese medicine. Along with the development of sensors and computer technology, the objectivity of the Chinese medicine Pulse- feeling is necessary, which is the basis of this paper.Artificial neural network(ANN) is a class of interconnected network, which imitates the biological interconnections in structures. ANN has the strong self-adaptability and self-learning-ability as well as excellent robustness and tolerance ability. By the network, an optional nonlinear input-output mapping relationship can be realized. Concrete mapping relationship materializes at the distributed linking weight values between neurons that build up the ANN. A trained neural network may be used to realize pattern recognitions, signal processing and detection, etc..Considering the characteristic differences between the pulse signals of heroin addicts and healthy persons, we successfully use BP network and self-organizing network to identify heroin addicts from the pulse signals of 15 heroin addicts and 15 healthy persons. Firstly, a two-layer BP network with 40~20~1 is constructed in this paper. The input signals of the network are obtained by clipping the characteristic section of every pulse signal. The network is trained by the training samples obtained by the clipping and the training samples with additive noise, respectively. When the 30 original pulse signals and the ones with additive noise are tested in the trained network, all of the heroin addicts are correctly identified except that one healthy person Z01 is misjudged and the exactness ratio reaches to 96.7%. The training speed of the basic BP algorithm is compared with that of the Levenberg-Marquardt algorithm also. The experiment shows the pulse signals are identified except that healthy persons Z01 and Z10 and heroin addict B13 are misjudged when the 30 pulse signals are analysed in the self-organizing network. It is also found that the network is sensitive to noise and the result is inferior to that of the network without noise when 30 samples with additive noise is analysed by the self-organizing network.The research shows that the exactness ratio is high, the tolerance ability is excellent and the training speed is slow when the BP network analyses the pulse signals. And the research also shows that the training speed is fast, the exactness ratio and tolerance ability is inferior to that of the BP network. Besides the theories and algorithms of the BP network and the self-organizing network are deduced and proved in this paper, and the basic theories coherent to neural network are collected and generalized, which includes the classification of the neural network structure, the algorithms and the training function.