论文部分内容阅读
由于心电(ECG)信号特征参数具有模糊性和随机性的特点,为ECG信号的自动分析增加了难度,使ECG自动分析往往很难完全达到专家诊断的效果。针对这一问题,本文探索性地将集随机性和模糊性于一体的定性定量不确定性转化模型——云模型用于ECG信号自动分析系统中。基于云变换和综合云的思想,实现ECG信号的聚类分析。进一步基于云模型理论描述心电专家在临床总结出的分类规则,进而分析诊断,从而克服了传统ECG自动分析诊断过程中判断指标、阈值的绝对化和判断规则的精确化的问题。与传统方法相比,分析过程更加接近于心电专家的思维与逻辑,分析结果更加符合医学专家的诊断结果。以MIT/BIH心电数据库为对象,实验结果表明,结果更加接近于心电专家的模糊逻辑思维分析的结果,是一种有效的ECG信号分析方法。
Due to the fuzziness and randomness of ECG signal characteristic parameters, the automatic analysis of ECG signals increases the difficulty. It is often difficult for ECG automatic analysis to fully achieve the effect of expert diagnosis. In response to this problem, this paper explores qualitatively quantitative uncertain transformation model integrating cloudiness and randomness into an automatic ECG signal analysis system. Clustering analysis of ECG signals based on the idea of cloud transformation and integrated cloud. Furthermore, based on the cloud model theory, the classification rules summarized by clinicians in ECG are further described, and the diagnosis is further analyzed. This overcomes the problems of the traditional methods such as the judgment of indicators, the absolute threshold and the accuracy of judgment rules. Compared with traditional methods, the analysis process is closer to the thinking and logic of ECG experts, and the analysis results are more in line with the diagnosis results of medical experts. Taking MIT / BIH ECG database as an example, the experimental results show that the result is closer to the result of fuzzy logical thinking analysis of ECG experts and is an effective ECG signal analysis method.