胎儿心电信号检测算法的研究进展  

Advances in fetal electrocardiogram detection algorithms

作　　者：郝婧宇 南格丽 吴水才[1] HAO Jing-yu;NAN Ge-li;WU Shui-cai(Department of Biomedical Engineering,Faculty of Environmental and Life Sciences,Beijing University of Technology,Beijing 100124,China)

机构地区：[1]北京工业大学环境与生命学部生物医学工程系,北京100124

出　　处：《医疗卫生装备》2021年第6期90-96,102,共8页Chinese Medical Equipment Journal

摘　　要：介绍了胎儿心电信号(fetal electrocardiogram,FECG)的预处理方法,综述了差分阈值法(difference threshold arithmetic,DTA)、小波变换(wavelet transform,WT)法、经验模态分解(empirical mode decomposition,EMD)法、机器学习算法和多导联融合检测算法等常用的FECG检测算法的原理及其国内外研究现状,分析了各FECG检测算法的优缺点,指出了融合多种检测算法、设计轻量型训练模型、引入迁移学习算法和提高算法自学习能力是未来FECG检测算法的改进方向。The preprocessing methods of fetal electrocardiogram(FECG)were introduced,and the principles of common FECG detection algorithms were reviewed such as difference threshold arithmetic(DTA),wavelet transform(WT)algorithm,empirical mode decomposition(EMD)algorithm,machine learning algorithm and multi-lead fusion detection algorithm.The advantages and disadvantages of each FECG detection algorithm were analyzed,and it's pointed out that fusing multiple detection algorithms,designing lightweight training models,introducing migration learning algorithms and improving self-learning capability of the algorithms were the key factors for future FECG detection.

关 键 词：胎儿心电信号 差分阈值法 小波变换法 经验模态分解法 机器学习算法 多导联融合检测算法 

分 类 号：R318[医药卫生—生物医学工程] R540.41[医药卫生—基础医学]