基于模糊逻辑推理和神经网络的呼吸障碍诊断方法研究
[Study on diagnostic methods of breathing disorders based on fuzzy logic inference and the neural network].
作者信息
Chen Min, Yin Xuezhi
机构信息
Shanghai Medical Industry College, University of Shanghai for Science and Technology, Shanghai 200093, China.
出版信息
Zhongguo Yi Liao Qi Xie Za Zhi. 2011 Jul;35(4):260-2.
This paper descries a new non-invasive method for diagnosis of breathing disorders based on adaptive-network-based fuzzy inference system (ANFIS). In this method, PetCO2, SpO2 and HR are chosen as inputs, and the breathing condition is selected as output ofANFIS. The inputs and output are then classified into fuzzy subsets by experts' knowledge. After, the fuzzy IF-THEN rules are built up according to the corresponding membership functions by set up of fuzzy subsets. The neural network was finally established and the membership functions and fuzzy rules were optimized by training. The results of experiment shows that ANFIS is more effective than BP Network regarding the diagnosis of breathing disorders.
本文描述了一种基于自适应网络模糊推理系统（ANFIS）的用于诊断呼吸障碍的新型非侵入性方法。在该方法中，选择呼气末二氧化碳分压（PetCO2）、血氧饱和度（SpO2）和心率（HR）作为输入，呼吸状况作为ANFIS的输出。然后根据专家知识将输入和输出分类为模糊子集。之后，通过建立模糊子集，根据相应的隶属函数建立模糊IF-THEN规则。最终建立神经网络，并通过训练对隶属函数和模糊规则进行优化。实验结果表明，在呼吸障碍诊断方面，ANFIS比BP网络更有效。
Zotero 插件