一种基于递归分析的宫缩信号强度分类方法  

A Classification Method of Uterine Contraction Signal Strength Based on Recursive Analysis

作　　者：刘志康 张烨菲 邵李焕[1] 张钰[1] LIU Zhikang;ZHANG Yefei;SHAO Lihuan;ZHANG Yu(School of Electronic Information,Hangzhou Dianzi University,Hangzhou Zhejiang 310018,China)

机构地区：[1]杭州电子科技大学电子信息学院,浙江杭州310018

出　　处：《杭州电子科技大学学报（自然科学版）》2020年第5期39-45,共7页Journal of Hangzhou Dianzi University：Natural Sciences

基　　金：浙江省基础公益研究计划资助项目(LGG19F010010)。

摘　　要：提出一种基于递归分析的宫缩信号分类方法,用于区分不同强度的宫缩信号。首先,使用经验模态分解(Empirical Mode Decomposition,EMD)与形态学结合的联合滤波算法进行高频降噪,并通过平滑先验法(Smoothness Priors Approach,SPA)去除信号的基线漂移;然后,通过递归分析获取描述递归图的递归参量,将其与信号的一维时域、形态特征融合成16维的特征向量;最后,使用SMOTE-PCA-SVM分类器对宫缩信号强度进行分类。仿真实验表明:所提宫缩信号强度分类方法与传统方法相比,在准确性、灵敏度和特异性方面均有提高,分别达到98.22%,98.08%以及98.17%。This paper proposes a classification method of uterine contraction signals based on recursive analysis,which is used to distinguish different strength of uterine contraction signals.Firstly,the federated filtering algorithm of Empirical Mode Decomposition(EMD)and Morphology is used for high frequency noise reduction,and the baseline drift of the signal is removed by the Smooth Priors Approach(SPA).Secondly,the recursive parameters describing the recursive graph are obtained by recursive analysis,which are fused with the one-dimensional time-domain and morphological features of the signal into 16 dimensional eigenvectors.Finally,SMOTE-PCA-SVM classifier is used to classify the strength of contractions.The simulation results show that the accuracy,sensitivity and specificity of the proposed method can reach 98.22%,98.08%and 98.17%respectively,which are better than those of the traditional algorithms.

关 键 词：宫缩强度 经验模态分解 递归分析 SMOTE-PCA-SVM分类器 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]