心电图的智能识别技术

Intelligent Recognition Technology of ECG

模糊逻辑、神经网络是人工智能的重要分支，它们从不同角度、在一定程度上模拟了人类智能。本文先后将模糊逻辑、神经网络以及模糊神经网络技术用于心电图识别，获得了良好的效果。在模糊识别方面，从模糊识别矩阵的建立到模糊输入向量的确定，是针对此类具体问题的多传感器模糊信息融合算法，既综合考虑了各输入变量的作用，又突出了识别的主要依据。本文还给出了神经网络识别的三种试验结果及其与模糊神经网络识别的对比。模糊神经网络既充分发挥了神经网络的学习功能，又充分发挥了模糊逻辑的推理功能，因此具有很高的识别精度。

Fuzzy logic and neural network are both important branches of artificial intelligence. From some degree they simulate human intelligence in different aspects. In this paper fuzzy logic, neural network and fuzzy neural network are applied to ECG recognition respectively; and the results are very good. In the aspect of fuzzy recognition, from the establishment of fuzzy classification matrix to the determination of fuzzy input vector, it is the multisensor fuzzy information fusion algorithm of the discrete problems; thus not only the interaction of input variables is considered but also the main evidence of patterns is focused. Three different experimented results of neural network and the comparison with those of fuzzy neural network are given. Fuzzy neural network can use both the learning function of neural network and the reasoning function of fuzzy logic, so it has high accuracy of identification.