摘要
随着机器学习的飞速发展,不少研究人员将智能分类算法应用于胎儿健康状态评估中。本文首先介绍了用于胎儿健康状态评估的主要特征参数,综述了基于特征提取与基于深度学习的智能分类方法的国内外研究进展,主要包括决策树与随机森林、最小二乘向量机、K最近邻算法、卷积神经网络、循环神经网络等算法,分析了这些方法在胎儿健康状态智能分类中的优势和存在的问题,最后对其进行总结与展望。
In recent years,with the rapid development of machine learning,many researchers apply intelligent learning algorithms to the intelligent evaluation of fetal state,this paper introduces the criteria and extraction methods of fetal state evaluation characteristic parameters,summarizes the progress of domestic and foreign research based on feature extraction and intelligent classification methods based on deep learning,including decision trees and random forests,vector machines and least-2 vector machines,K near neighbor algorithms,reel neural networks,circulatory neural networks and other algorithms.This paper analyzes the advantages and problems of these methods in the intelligent classification of fetal state,and finally summarizes and looks forward to them.
作者
郝婧宇
南格丽
吴水才
HAO Jingyu;NA Geli;WU Shuicai(Department of Biomedical Engineering,Beijing University of technology,Bering 100000,China)
出处
《生命科学仪器》
2021年第3期29-37,共9页
Life Science Instruments
关键词
胎儿健康状态评估
智能分类
机器学习
卷积神经网络
循环神经网络
fetal status assessment
smart classification
machine learning
cogeneration neural networks
circulatory neural networks