Ecg Parameters Of The Neural Network Marked Its Applications In Speciation Analysis

Parameters and shapes of dynamic Electrocardiograph are main reference in diagnosing various heart diseases. However, in early times, parameter labels and shape analysis are manually implemented by doctors, in addition, those data are large. These limitations debase diagnosing reliability. Since neural networks are parallel distributive processors based on human thinking, it is available to achieve this function using neural networks.Based on the research of Fuzzy ARTMAP, simplified Fuzzy ARTMAP is used in the field of parameter labels and shape analysis in this paper. And the Electrocardiograph classify has been achieve thought the simplified Fuzzy ARTMAP classifier.The main contribution and innovation of this dissertation are showed as follows:1. The profound dynamical characteristic analysis and mechanism analysis of fuzzy ARTMAP classifier is carried through. The analysis includes the choice of some parameters. And geometric interpretation of fuzzy ART classifier is also presented.2. The profound characteristic analysis of simplified and modified Fuzzy ARTMAP classifier is carried through. The structure of fuzzy ARTMAP classifier is too complex if it is just used to simple classify. Base on the characteristic of cardiogram, simplified and modified Fuzzy ARTMAP classifier has been used in classify on the ST segment of cardiogram. The modified algorithm only uses one membership function to calculate the similarity between the input vector and the weight of winner neuron. The feasibility of the SFAM neural networks has been validated by two simulated experiments at the end of paper.3. Thought extract the feature of the cardiogram for verification by use the different methods, both the simplified classifier and the classic classifier can classify the cardiogram automatically. At the same time the simplified classifier with smaller spending time.