论文部分内容阅读
随着心电图数据量快速增长,计算机辅助心电图分析也有着越来越广阔的应用需求。本文在基于导联卷积神经网络的临床心电图分类算法上提出多种策略,进一步提升其在实际应用中的性能。首先用不同的预处理方法和训练方法获得两个不同的分类器,接着用多重输出预测法来增强每个分类器的性能,最后用贝叶斯方法进行融合。测试了超过15万条心电图记录,所提方法的准确率和受试者工作特征曲线下面积(AUC)分别为85.04%和0.918 5,明显优于基于特征提取的传统方法。
With the rapid growth of ECG data, computer-aided ECG analysis also has more and more broad application needs. In this paper, a variety of strategies are proposed in the clinical ECG classification algorithm based on the convolution neural network of the lead to further enhance its performance in practical applications. Firstly, two different classifiers are obtained by different preprocessing methods and training methods, then the multiple output prediction method is used to enhance the performance of each classifier. Finally, Bayesian method is used to fuse the classifiers. More than 150,000 ECG records were tested. The accuracy of the proposed method and the area under the receiver operating characteristic curve (AUC) were 85.04% and 0.918 5, respectively, which was significantly better than the traditional method based on feature extraction.