Abstract:

At present, some rule-based or machine-based learning methods have been proposed to automatically classify the input Electrocardiogram（ECG） signal as clinically acceptable or unacceptable, and achieved good results. In this paper, an ECG signal quality assessment method based on convolution neural network is proposed, in which the features can be automatically learned by network model. Moreover, the strategy that evaluating fragment by fragment is adopted, which is different with other methods to evaluate the quality from a global perspective. In order to compare with other methods, local evaluating results together to get the overall quality assessment of ECGs are combined, the test results on public data show the good performance of the method.

Key words: electrocardiogram, signal quality analysis, convolution neural network

摘要：

目前，已有一些研究者提出基于规则或基于机器学习的方法，自动将输入的心电信号识别为临床可接受或不可接受，取得了不错的效果。提出基于卷积神经网络的心电信号质量评估方法，通过网络模型自动学习分类特征，减少人工干预。同时，与其他方法从全局角度对心电信号质量进行评估的策略不同的是，采用局部评估方法，既可对心电图的部分片段数据进行质量分析，也可通过简单融合策略对全局信号的质量进行评估。在公开数据集上的测试结果表明，提出的方法比其他几个对比方法的判断准确度高。

关键词: 心电图, 信号质量评估, 卷积神经网络