【摘要】 为了解决在阵发性房颤中持续时长较短的房颤检测问题,本文提出了一种基于随机森林的阵发性房颤自动检测算法。该算法主要是利用了在房颤发生时期心电图的RR间期绝对不规则这一特点,并采用了一种滑动窗口模式,分别从时域、频域和非线性分析3个方面进行特征提取,最后通过随机森林进行房颤检测,并将检测结果与数据实际标注对比。数据采用的是MIT-BIH房颤数据库,其最终敏感性(Se)、特异性(Sp)和阳性检测率(+P)分别为99.12%、98.86%、98.46%,实验结果证明了该算法的有效性。更多还原

【Abstract】 In order to solve the problem of short duration detection in paroxysmal atrial fibrillation,an automatic detection method was proposed based on random forest in this paper.This algorithm mainly used the absolute irregular of electrocardiogram in the periods between the RR and adopted a sliding window model.The features were extracted from the time domain,the frequency domain and the nonlinear analysis.Finally,atrial fibrillation was detected by random forest,and the results were compared with the actual data marking.The data was from the MIT-BIH database,where the sensitivity,specificity and positive detection rates were 99.12%,98.86% and 98.46%,respectively.The experimental results showed the effectiveness of the proposed algorithm.更多还原