EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Applying Recurrent Neural Networks for Anomaly Detection in Electrocardiogram Sensor Data

Ana Minic iD, Luka Jovanovic iD, Nebojsa Bacanin iD, Catalin Stoean iD, Miodrag Zivkovic iD, Petar Spalevic iD, Aleksandar Petrovic iD, Milos Dobrojevic iD, Ruxandra Stoean iD

Sensors · 17.12.2023 · 23 (24) · 9878

DOI: 10.3390/s23249878 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 17.12.2023. Verfasst von Ana Minic und 8 weiteren Autor:innen. Beteiligte Einrichtung: Teacher Education Faculty, University of Pristina in Kosovska Mitrovica, 38220 Kosovska Mitrovica, Serbia. Hauptthemen laut MEDLINE: Neural Networks, Computer, Algorithms. Die Arbeit wurde bisher 59-mal zitiert (Crossref) und nennt 70 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: COST Action ROAR-NET - Randomised Optimisation Algorithms Research Network, COST (European Cooperation in Science and Technology).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Monitoring heart electrical activity is an effective way of detecting existing and developing conditions. This is usually performed as a non-invasive test using a network of up to 12 sensors (electrodes) on the chest and limbs to create an electrocardiogram (ECG). By visually observing these readings, experienced professionals can make accurate diagnoses and, if needed, request further testing. However, the training and experience needed to make accurate diagnoses are significant. This work explores the potential of recurrent neural networks for anomaly detection in ECG readings. Furthermore, to attain the best possible performance for these networks, training parameters, and network architectures are optimized using a modified version of the well-established particle swarm optimization algorithm. The performance of the optimized models is compared to models created by other contemporary optimizers, and the results show significant potential for real-world applications. Further analyses are carried out on the best-performing models to determine feature importance.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Neural Networks, ComputerAlgorithmsElectrocardiography

Stichwörter der Autor:innen

diagnosiselectrocardiogramoptimizationparticle swarm optimizationrecurrent neural networks

Autor:innen und Institutionen

  • Teacher Education Faculty, University of Pristina in Kosovska Mitrovica, 38220 Kosovska Mitrovica, Serbia.
  • Faculty of Informatics and Computing, Singidunum University, 160622 Belgrade, Serbia.
  • Department of Computer Science, Faculty of Sciences, University of Craiova, 200585 Craiova, Romania.
  • Faculty of Technical Science, University of Pristina in Kosovska Mitrovica, Filipa Visnjica bb, 38220 Kosovska Mitrovica, Serbia.

ORCID-Profile: Ana Minic, Luka Jovanovic, Nebojsa Bacanin, Catalin Stoean, Miodrag Zivkovic, Petar Spalevic, Aleksandar Petrovic, Milos Dobrojevic, Ruxandra Stoean

Förderung

  • COST Action ROAR-NET - Randomised Optimisation Algorithms Research Network
  • COST (European Cooperation in Science and Technology)

Zitieren

APA

Minic, A., Jovanovic, L., Bacanin, N., Stoean, C., Zivkovic, M., Spalevic, P., Petrovic, A., Dobrojevic, M., & Stoean, R. (2023). Applying Recurrent Neural Networks for Anomaly Detection in Electrocardiogram Sensor Data. Sensors, 23(24), 9878. https://doi.org/10.3390/s23249878

Vancouver

Minic A, Jovanovic L, Bacanin N, Stoean C, Zivkovic M, Spalevic P, et al. Applying Recurrent Neural Networks for Anomaly Detection in Electrocardiogram Sensor Data. Sensors. 2023;23(24):9878. doi:10.3390/s23249878

BibTeX
@article{Minic2023,
  author = {Minic, Ana and Jovanovic, Luka and Bacanin, Nebojsa and Stoean, Catalin and Zivkovic, Miodrag and Spalevic, Petar and Petrovic, Aleksandar and Dobrojevic, Milos and Stoean, Ruxandra},
  title = {Applying Recurrent Neural Networks for Anomaly Detection in Electrocardiogram Sensor Data},
  journal = {Sensors},
  year = {2023},
  volume = {23},
  number = {24},
  pages = {9878},
  doi = {10.3390/s23249878}
}
Kostenlos starten →