人工智能预测心房颤动研究
Artificial Intelligence Prediction of Atrial Fibrillation
阿力木江·图尔荪 1肖慧 2张鑫 2林铭俊 2陈超敏 2洪永2
作者信息
- 1. 南方医科大学中西医结合医院,广州 510515
- 2. 南方医科大学生物医学工程学院,广州 510515
- 折叠
摘要
探讨人工智能分析心电图预测心房颤动发作风险的有效性.人工智能应用于心房颤动的预测方法有基于早发性心房复合体的检测方法、基于P波形态学的分析方法、基于心率变异性(HRV)的分析方法和基于深度学习进行分析的方法.结果显示,基于HRV分析方法的房颤预测准确率达98.21％,基于深度学习进行房颤预测的准确率达93.56％.应继续优化基于HRV特征的分析方法以及将一些新技术引入房颤预测领域,以期达到更好的预测效果.
Abstract
The study investigates the effectiveness of ECG analysis by artificial intelligence in predicting the risk of atrial fibrillation.Four main methods are introduced:the detection method based on premature atrial complexity,the analysis method based on P-wave morphology,the analysis method based on heart rate variability,and the method based on deep learning for analysis.The results show that the accuracy rate of atrial fibrillation prediction based on HRV analysis method is 98.21％,and the accuracy rate of atrial fibrillation prediction based on deep learning is 93.56％.It is necessary to continuously optimize the analysis methods based on HRV characteristics,and introduce some new techniques into atrial fibrillation prediction in order to achieve better prediction results.
关键词
人工智能/心房颤动/房颤预测/心率变异性Key words
Artificial intelligence/Atrial fibrillation/Atrial fibrillation/Heart rate variability引用本文复制引用
基金项目
国家重点研发计划项目(2023YFC2414500)
出版年
2024
NETL
NSTL
万方数据