Analyzing artificial intelligence systems for the prediction of atrial fibrillation from sinus-rhythm ECGs including demographics and feature visualization
Entity
UAM. Departamento de Tecnología Electrónica y de las ComunicacionesPublisher
Nature PortfolioDate
2021-11-23Citation
10.1038/s41598-021-02179-1
Scientific Reports 11.1 (2021): 22786
ISSN
2045-2322 (online)DOI
10.1038/s41598-021-02179-1Funded by
GJO, AS-G, LJJ-B received a research grant from the Carlos III Institute of Health under the health Strategy action 2020-2022 with reference PI20/00792. Tis study is also supported partially by projects TRESPASS-ETN (H2020-MSCAITN-2019-860813), PRIMA (H2020-MSCA-ITN-2019-860315), IDEA-FAST (IMI2-2018-15-853981), BIBECA (RTI2018-101248-B-I00 MINECO/FEDER)Project
info:eu-repo/H2020-MSCAITN-2019-860813; info:eu-repo/H2020-MSCA-ITN-2019-860315; info:eu-repo/IMI2-2018-15-853981; Gobierno de España. RTI2018-101248-B-I00Editor's Version
https://doi.org/10.1038/s41598-021-02179-1Subjects
TelecomunicacionesRights
© The author(s)Abstract
Atrial fibrillation (AF) is an abnormal heart rhythm, asymptomatic in many cases, that causes several health problems and mortality in population. This retrospective study evaluates the ability of different AI-based models to predict future episodes of AF from electrocardiograms (ECGs) recorded during normal sinus rhythm. Patients are divided into two classes according to AF occurrence or sinus rhythm permanence along their several ECGs registry. In the constrained scenario of balancing the age distributions between classes, our best AI model predicts future episodes of AF with area under the curve (AUC) 0.79 (0.72–0.86). Multiple scenarios and age-sex-specific groups of patients are considered, achieving best performance of prediction for males older than 70 years. These results point out the importance of considering different demographic groups in the analysis of AF prediction, showing considerable performance gaps among them. In addition to the demographic analysis, we apply feature visualization techniques to identify the most important portions of the ECG signals in the task of AF prediction, improving this way the interpretability and understanding of the AI models. These results and the simplicity of recording ECGs during check-ups add feasibility to clinical applications of AI-based models
Files in this item
Google Scholar:Melzi, Pietro
-
Tolosana Moranchel, Rubén
-
Cecconi, Alberto
-
Sanz-Garcia, Ancor
-
Ortega, Guillermo J.
-
Jimenez-Borreguero, Luis Jesus
-
Vera Rodríguez, Rubén
This item appears in the following Collection(s)
Related items
Showing items related by title, author, creator and subject.