Prediction of Atrial Fibrillation:From Traditional Regression Analysis Towards Artificial Intelligence

VernacularTitle:心房颤动预测:从传统回归模型分析到人工智能模型
Author: Xiaoqing ZHU ¹ ; Tao CHEN ; Juan SHEN ; Jun GUO
Author Information

1. 中国人民解放军总医院第六医学中心心血管病医学部北京 100048;中国人民解放军医学院北京 100853
Keywords: atrial fibrillation; artificial intelligence; machine learning; electrocardiogram; radiomics; prediction
From: Chinese Circulation Journal 2023;38(12):1305-1310
CountryChina
Language:Chinese
Abstract: Atrial fibrillation,which is closely correlated with stroke and heart failure,is among the most common arrhythmias in clinical practice.The asymptomatic and paroxysmal nature of atrial fibrillation leads to a very high missed diagnosis rate,more than half of misdiagnosed patients are classified as intermediate to high risk population of ischemic stroke and heart failure.With the progress of population aging in China,the prevalence of atrial fibrillation will continue to increase.The prediction model of atrial fibrillation can help screen patients with high risk of atrial fibrillation.Subsequent preventive intervention and intensive monitoring for this population can reduce the incidence of stroke and heart failure.This paper summarizes the research progress of traditional regression model and artificial intelligence in atrial fibrillation prediction,and limitations of current research and future research directions.