Pulse Recognition Neural Network Integrated Circuits Research

The artificial neural network that used to differentiate different patterns based on the application of the pregnant woman's pulse recognition is put forward in this paper. And then, we studied and discussed the implement of neural network integrated circuit.Based on artificial neural network theory, a neural network is designed to determine pregnancy by taking the pulses adopting the error back-propagation network (BP). The sample data are firstly analyzed and processed to form a sample database of the network. Then, the network is trained and emulated. And the emulation result is further analyzed. Building upon make an in-depth analysis, study on the plan of the neuron integrated circuit emphatically.The emulation result show that the rate of accuracy of this neural network system differentiating the pregnant woman's pulse is up to 100% . In spite of limited training samples, the experiment and analysis indicate the artificial neural network method has obvious superiority and extensive prospect of application in the field of pulse-taking diagnosis in Chinese medicine.