ECG Restitution Analysis and Machine Learning to Detect Paroxysmal Atrial Fibrillation: Insight from the Equine Athlete as a Model for Human Athletes.
View / Open Files
Authors
Huang, Ying H
Alexeenko, Vadim
Tse, Gary
Huang, Christopher L-H
Marr, Celia M
Publication Date
2021Journal Title
Function (Oxf)
ISSN
2633-8823
Publisher
Oxford University Press (OUP)
Language
eng
Type
Article
This Version
VoR
Metadata
Show full item recordCitation
Huang, Y. H., Alexeenko, V., Tse, G., Huang, C. L., Marr, C. M., & Jeevaratnam, K. (2021). ECG Restitution Analysis and Machine Learning to Detect Paroxysmal Atrial Fibrillation: Insight from the Equine Athlete as a Model for Human Athletes.. Function (Oxf) https://doi.org/10.1093/function/zqaa031
Abstract
Atrial fibrillation is the most frequent arrhythmia in both equine and human athletes. Currently, this condition is diagnosed via electrocardiogram (ECG) monitoring which lacks sensitivity in about half of cases when it presents in paroxysmal form. We investigated whether the arrhythmogenic substrate present between the episodes of paroxysmal atrial fibrillation (PAF) can be detected using restitution analysis of normal sinus-rhythm ECGs. In this work, ECG recordings were obtained during routine clinical work from control and horses with PAF. The extracted QT, TQ, and RR intervals were used for ECG restitution analysis. The restitution data were trained and tested using k-nearest neighbor (k-NN) algorithm with various values of neighbors k to derive a discrimination tool. A combination of QT, RR, and TQ intervals was used to analyze the relationship between these intervals and their effects on PAF. A simple majority vote on individual record (one beat) classifications was used to determine the final classification. The k-NN classifiers using two-interval measures were able to predict the diagnosis of PAF with area under the receiving operating characteristic curve close to 0.8 (RR, TQ with k ≥ 9) and 0.9 (RR, QT with k ≥ 21 or TQ, QT with k ≥ 25). By simultaneously using all three intervals for each beat and a majority vote, mean area under the curves of 0.9 were obtained for all tested k-values (3-41). We concluded that 3D ECG restitution analysis can potentially be used as a metric of an automated method for screening of PAF.
Keywords
ECG, Paroxysmal atrial fibrillation, Diagnostic, Machine Learning, Equine Athletes, Restitution Analysis
Sponsorship
PetPlan Charitable Trust (S17-447-485)
Identifiers
35330977, PMC8788737
External DOI: https://doi.org/10.1093/function/zqaa031
This record's URL: https://www.repository.cam.ac.uk/handle/1810/336441
Statistics
Total file downloads (since January 2020). For more information on metrics see the
IRUS guide.
Recommended or similar items
The current recommendation prototype on the Apollo Repository will be turned off on 03 February 2023. Although the pilot has been fruitful for both parties, the service provider IKVA is focusing on horizon scanning products and so the recommender service can no longer be supported. We recognise the importance of recommender services in supporting research discovery and are evaluating offerings from other service providers. If you would like to offer feedback on this decision please contact us on: support@repository.cam.ac.uk