基于卷积神经网络与通道和空间注意力机制的房颤预测模型研究

Study on atrial fibrillation prediction model based on convolutional neural network and CBAM attention mechanism

目的采用人工智能技术提出一种模型,以对房颤进行早期预防和诊断。方法提出一种基于卷积神经网络(convolutional neural network,CNN)与通道和空间注意力机制(convolutional block attention module,CBAM)的模型用于对房颤的诊断与预测。结果根据长期心房颤动数据库、MIT-BIH心房颤动数据库和MIT-BIH正常窦性心律数据库的数据,提出的模型在全盲的情况下总体准确率达94.2%。结论提出的模型满足了医学心电图解释的需要,为房颤的预测研究提供了新思路。

Objective To propose a model for early prevention and diagnosis of atrial fibillation by using artificial intelli-gence technology.Methods A model based on convolutional neural network(CNN)and convolutional block attention module(CBAM)was proposed for the diagnosis and prediction of atrial fibrillation.Results The overall accuracy of the proposed model reached 94.2%in the case of total blindness based on the data from the long term atrial fibrllation database,the MIT-BIH atrial fibrillation database and the MIT-BIH normal sinus rhythm database.Conclusion The proposed method satisfies the needs of medical ECG interpretation and provides a new idea for the prediction of atrial fibrillation.