论文部分内容阅读
为了解决传统的房颤检测算法中P波形态多变而不易提取特征的问题,本文提出了一种基于多特征融合与卷积神经网络结合的房颤检测算法。首先,分别提取单心拍心房活动信号递归矩阵的特征值及相邻两个心拍的心房活动信号的相干谱来得到底层特征;然后,分别采用卷积神经网络对底层特征进行分析;最后,采用决策级融合来改善算法的性能。经MIT-BIH房颤数据库验证,该算法的正确率,灵敏度,特异性分别可达95.62%,99.88%,91.36%。结果表明,该方法能有效解决特征提取困难,泛化能力差的问题。
In order to solve the problem that the traditional Atrial Fibrillation (AF) detection algorithm is not easy to extract features due to the change of P wave morphology, this paper proposes a AF detection algorithm based on multi-feature fusion and convolutional neural network. Firstly, the eigenvalues of recursive matrices of single-heart beat atrial signals and the coherence spectra of atrial activity signals of two adjacent echocardiograms were extracted respectively to obtain the underlying features. Then, the convolution neural network was used to analyze the underlying features respectively. Finally, Level fusion to improve the performance of the algorithm. The MIT-BIH atrial fibrillation database verified that the accuracy, sensitivity and specificity of the algorithm were up to 95.62%, 99.88% and 91.36% respectively. The results show that this method can effectively solve the problem of feature extraction and poor generalization ability.