A Deep Learning Framework Based on Denoising and 2D Image Encoding for Arrhythmia Classification
Sensors · 16.08.2026 · 26 (16) · 5183
DOI: 10.3390/s26165183 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 16.08.2026. Verfasst von Ji-Yun Seo und 2 weiteren Autor:innen. Beteiligte Einrichtung: AI Education Center, Soonchunhyang University, Asan 31518, Republic of Korea. Hauptthemen laut MEDLINE: Arrhythmias, Cardiac, Deep Learning, Image Processing, Computer-Assisted. Die Arbeit nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Soonchunhyang University.
Abstract (Originalsprache)
Electrocardiogram (ECG) signals are essential for arrhythmia detection; however, they are frequently degraded by noise during acquisition, and their evaluation is vulnerable to data-leakage and patient-overlap issues that can compromise model assessment. Therefore, in this study, we propose an image-encoding-based arrhythmia classifier combined with a morphology-aware denoising autoencoder. We evaluate signals under a corrected, patient-independent protocol in which every model-selection decision was made on a separate validation partition. On a leakage-free test partition, the autoencoder achieved an SNR improvement of 5.63 dB and a correlation coefficient of 0.801, improving R-peak amplitude preservation and redetection accuracy under moderate-to-severe noise while introducing measurable morphology degradation when the input was already lightly contaminated. The proposed model encodes the denoised, beat-centered signal into images through an interleaved-grouping outer product with sorting and flipping, and classifies them with a multi-scale three-dimensional convolutional network. Under this protocol, the proposed model obtained the highest macro-F1 among five image-encoding architectures, but did not outperform four models operating directly on the denoised signal (macro-F1 32.5% versus 37.3–40.0%), indicating that the proposed encoding does not improve overall five-class classification under rigorous inter-patient evaluation. A controlled test showed that the encoding is exactly invariant to global signal-polarity inversion, unlike the sequential models. This targeted invariance, rather than a general accuracy advantage, is the contribution reported here.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- AI Education Center, Soonchunhyang University, Asan 31518, Republic of Korea.
ORCID-Profile: Ji-Yun Seo, Byeong Ho Park, Chang Min Kim
Förderung
- Soonchunhyang University
Zitieren
APA
Seo, J. Y., Park, B. H., & Kim, C. M. (2026). A Deep Learning Framework Based on Denoising and 2D Image Encoding for Arrhythmia Classification. Sensors, 26(16), 5183. https://doi.org/10.3390/s26165183
Vancouver
Seo JY, Park BH, Kim CM. A Deep Learning Framework Based on Denoising and 2D Image Encoding for Arrhythmia Classification. Sensors. 2026;26(16):5183. doi:10.3390/s26165183
BibTeX
@article{Seo2026,
author = {Seo, Ji-Yun and Park, Byeong Ho and Kim, Chang Min},
title = {A Deep Learning Framework Based on Denoising and 2D Image Encoding for Arrhythmia Classification},
journal = {Sensors},
year = {2026},
volume = {26},
number = {16},
pages = {5183},
doi = {10.3390/s26165183}
}
EUVIMED