Robust screening of atrial fibrillation with distribution classification

Abstract Atrial fibrillation (AF) correlates with an increased risk of all-cause mortality or stroke, mainly due to undiagnosed patients and undertreatment. Its screening is thus a key challenge, for which machine learning methods hold the promise of cheaper and faster campaigns. The robustness of s...

Full description

Saved in:
Bibliographic Details
Main Authors: Pierre-François Massiani, Lukas Haverbeck, Claas Thesing, Friedrich Solowjow, Marlo Verket, Matthias Daniel Zink, Katharina Schütt, Dirk Müller-Wieland, Nikolaus Marx, Sebastian Trimpe
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-10090-2
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849332991251185664
author Pierre-François Massiani
Lukas Haverbeck
Claas Thesing
Friedrich Solowjow
Marlo Verket
Matthias Daniel Zink
Katharina Schütt
Dirk Müller-Wieland
Nikolaus Marx
Sebastian Trimpe
author_facet Pierre-François Massiani
Lukas Haverbeck
Claas Thesing
Friedrich Solowjow
Marlo Verket
Matthias Daniel Zink
Katharina Schütt
Dirk Müller-Wieland
Nikolaus Marx
Sebastian Trimpe
author_sort Pierre-François Massiani
collection DOAJ
description Abstract Atrial fibrillation (AF) correlates with an increased risk of all-cause mortality or stroke, mainly due to undiagnosed patients and undertreatment. Its screening is thus a key challenge, for which machine learning methods hold the promise of cheaper and faster campaigns. The robustness of such methods to varying artifacts, noise, and conditions is then crucial. We introduce the first distributional support vector machine (SVM) for robust detection of AF from short, noisy electrocardiograms. It achieves state-of-the-art performance and unprecedented robustness on the screening problem while only leveraging one interpretable feature and little training data. We illustrate these advantages by evaluating on other data sources (cross-data-set) and through sensitivity studies. These strengths result from two main components: (i) preliminary peak detection enabling robust computation of medically relevant features; and (ii) a mathematically principled way of aggregating those features to compare their full distributions. This establishes our algorithm as a relevant candidate for screening campaigns.
format Article
id doaj-art-bd36ba6a3fc744a9a1642e3f4d88cc1f
institution Kabale University
issn 2045-2322
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-bd36ba6a3fc744a9a1642e3f4d88cc1f2025-08-20T03:46:01ZengNature PortfolioScientific Reports2045-23222025-07-011511910.1038/s41598-025-10090-2Robust screening of atrial fibrillation with distribution classificationPierre-François Massiani0Lukas Haverbeck1Claas Thesing2Friedrich Solowjow3Marlo Verket4Matthias Daniel Zink5Katharina Schütt6Dirk Müller-Wieland7Nikolaus Marx8Sebastian Trimpe9Institute for Data Science in Mechanical Engineering, RWTH Aachen UniversityInstitute for Data Science in Mechanical Engineering, RWTH Aachen UniversityInstitute for Data Science in Mechanical Engineering, RWTH Aachen UniversityInstitute for Data Science in Mechanical Engineering, RWTH Aachen UniversityDepartment of Internal Medicine I, University Hospital RWTH AachenDepartment of Internal Medicine I, University Hospital RWTH AachenDepartment of Internal Medicine I, University Hospital RWTH AachenDepartment of Internal Medicine I, University Hospital RWTH AachenDepartment of Internal Medicine I, University Hospital RWTH AachenInstitute for Data Science in Mechanical Engineering, RWTH Aachen UniversityAbstract Atrial fibrillation (AF) correlates with an increased risk of all-cause mortality or stroke, mainly due to undiagnosed patients and undertreatment. Its screening is thus a key challenge, for which machine learning methods hold the promise of cheaper and faster campaigns. The robustness of such methods to varying artifacts, noise, and conditions is then crucial. We introduce the first distributional support vector machine (SVM) for robust detection of AF from short, noisy electrocardiograms. It achieves state-of-the-art performance and unprecedented robustness on the screening problem while only leveraging one interpretable feature and little training data. We illustrate these advantages by evaluating on other data sources (cross-data-set) and through sensitivity studies. These strengths result from two main components: (i) preliminary peak detection enabling robust computation of medically relevant features; and (ii) a mathematically principled way of aggregating those features to compare their full distributions. This establishes our algorithm as a relevant candidate for screening campaigns.https://doi.org/10.1038/s41598-025-10090-2Atrial fibrillationScreeningSupport vector machinesDistribution classificationMachine learning
spellingShingle Pierre-François Massiani
Lukas Haverbeck
Claas Thesing
Friedrich Solowjow
Marlo Verket
Matthias Daniel Zink
Katharina Schütt
Dirk Müller-Wieland
Nikolaus Marx
Sebastian Trimpe
Robust screening of atrial fibrillation with distribution classification
Scientific Reports
Atrial fibrillation
Screening
Support vector machines
Distribution classification
Machine learning
title Robust screening of atrial fibrillation with distribution classification
title_full Robust screening of atrial fibrillation with distribution classification
title_fullStr Robust screening of atrial fibrillation with distribution classification
title_full_unstemmed Robust screening of atrial fibrillation with distribution classification
title_short Robust screening of atrial fibrillation with distribution classification
title_sort robust screening of atrial fibrillation with distribution classification
topic Atrial fibrillation
Screening
Support vector machines
Distribution classification
Machine learning
url https://doi.org/10.1038/s41598-025-10090-2
work_keys_str_mv AT pierrefrancoismassiani robustscreeningofatrialfibrillationwithdistributionclassification
AT lukashaverbeck robustscreeningofatrialfibrillationwithdistributionclassification
AT claasthesing robustscreeningofatrialfibrillationwithdistributionclassification
AT friedrichsolowjow robustscreeningofatrialfibrillationwithdistributionclassification
AT marloverket robustscreeningofatrialfibrillationwithdistributionclassification
AT matthiasdanielzink robustscreeningofatrialfibrillationwithdistributionclassification
AT katharinaschutt robustscreeningofatrialfibrillationwithdistributionclassification
AT dirkmullerwieland robustscreeningofatrialfibrillationwithdistributionclassification
AT nikolausmarx robustscreeningofatrialfibrillationwithdistributionclassification
AT sebastiantrimpe robustscreeningofatrialfibrillationwithdistributionclassification