Machine learning for ranking f-wave extraction methods in single-lead ECGs
(2025) In Biomedical Signal Processing and Control 99.- Abstract
Introduction: The presence of fibrillatory waves (f-waves) is important in the diagnosis of atrial fibrillation (AF), which has motivated the development of methods for f-wave extraction. We propose a novel approach to benchmarking methods designed for single-lead ECG analysis, building on the hypothesis that better-performing AF classification using features computed from the extracted f-waves implies better-performing extraction. The approach is well-suited for processing large Holter data sets annotated with respect to the presence of AF. Methods: Three data sets with a total of 300 two- or three-lead Holter recordings, performed in the USA, Israel and Japan, were used as well as a simulated single-lead data set. Four existing... (More)
Introduction: The presence of fibrillatory waves (f-waves) is important in the diagnosis of atrial fibrillation (AF), which has motivated the development of methods for f-wave extraction. We propose a novel approach to benchmarking methods designed for single-lead ECG analysis, building on the hypothesis that better-performing AF classification using features computed from the extracted f-waves implies better-performing extraction. The approach is well-suited for processing large Holter data sets annotated with respect to the presence of AF. Methods: Three data sets with a total of 300 two- or three-lead Holter recordings, performed in the USA, Israel and Japan, were used as well as a simulated single-lead data set. Four existing extraction methods based on either average beat subtraction or principal component analysis (PCA) were evaluated. A random forest classifier was used for window-based AF classification. Performance was measured by the area under the receiver operating characteristic (AUROC). Results: The best performance was found for PCA-based extraction, resulting in AUROCs in the ranges 0.80–0.85, 0.66–0.80, and 0.87–0.92 for the data sets from USA, Israel, and Japan, respectively, when analyzed across leads; the AUROC of the simulated single-lead, noisy data set was 0.98. Conclusions: This study provides a novel approach to evaluating the performance of f-wave extraction methods, offering the advantage of not using ground truth f-waves for evaluation, thus being able to leverage real data sets for evaluation. The code is made open source at github.com/noambenmoshe/fwave.
(Less)
- author
- Ben-Moshe, Noam ; Brimer, Shany Biton ; Tsutsui, Kenta ; Suleiman, Mahmoud ; Sörnmo, Leif LU and Behar, Joachim A.
- organization
- publishing date
- 2025-01
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Atrial fibrillation, Biomedical signal processing, f-wave extraction, Machine learning, Performance evaluation
- in
- Biomedical Signal Processing and Control
- volume
- 99
- article number
- 106817
- publisher
- Elsevier
- external identifiers
-
- scopus:85203645280
- ISSN
- 1746-8094
- DOI
- 10.1016/j.bspc.2024.106817
- language
- English
- LU publication?
- yes
- id
- a371a35b-1792-468c-8824-447c67dbdc13
- date added to LUP
- 2024-11-12 10:43:24
- date last changed
- 2025-04-04 15:22:13
@article{a371a35b-1792-468c-8824-447c67dbdc13,
abstract = {{<p>Introduction: The presence of fibrillatory waves (f-waves) is important in the diagnosis of atrial fibrillation (AF), which has motivated the development of methods for f-wave extraction. We propose a novel approach to benchmarking methods designed for single-lead ECG analysis, building on the hypothesis that better-performing AF classification using features computed from the extracted f-waves implies better-performing extraction. The approach is well-suited for processing large Holter data sets annotated with respect to the presence of AF. Methods: Three data sets with a total of 300 two- or three-lead Holter recordings, performed in the USA, Israel and Japan, were used as well as a simulated single-lead data set. Four existing extraction methods based on either average beat subtraction or principal component analysis (PCA) were evaluated. A random forest classifier was used for window-based AF classification. Performance was measured by the area under the receiver operating characteristic (AUROC). Results: The best performance was found for PCA-based extraction, resulting in AUROCs in the ranges 0.80–0.85, 0.66–0.80, and 0.87–0.92 for the data sets from USA, Israel, and Japan, respectively, when analyzed across leads; the AUROC of the simulated single-lead, noisy data set was 0.98. Conclusions: This study provides a novel approach to evaluating the performance of f-wave extraction methods, offering the advantage of not using ground truth f-waves for evaluation, thus being able to leverage real data sets for evaluation. The code is made open source at github.com/noambenmoshe/fwave.</p>}},
author = {{Ben-Moshe, Noam and Brimer, Shany Biton and Tsutsui, Kenta and Suleiman, Mahmoud and Sörnmo, Leif and Behar, Joachim A.}},
issn = {{1746-8094}},
keywords = {{Atrial fibrillation; Biomedical signal processing; f-wave extraction; Machine learning; Performance evaluation}},
language = {{eng}},
publisher = {{Elsevier}},
series = {{Biomedical Signal Processing and Control}},
title = {{Machine learning for ranking f-wave extraction methods in single-lead ECGs}},
url = {{http://dx.doi.org/10.1016/j.bspc.2024.106817}},
doi = {{10.1016/j.bspc.2024.106817}},
volume = {{99}},
year = {{2025}},
}