EUVIMEDEUVIMED entdecken
Crossref

Channel Attention based 1D-CNN-LSTM Fully Connected Network for ECG Beat Classification using SE Block

V. Bhuvana Kruthi, L.V. Rajani Kumari, Y. Chalapathi Rao

International Journal of Drug Delivery Technology · 29.05.2026 · 16 (34s)

DOI: 10.25258/ijddt.16.34s.39 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Drug Delivery Technology, online erschienen am 29.05.2026. Verfasst von V. Bhuvana Kruthi und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

To automatically categorize ECG data for identifying cardiac arrhythmias, this thesis proposes a deep learning based technique. The proposed paradigm includes a convolutional neural network (CNN), long short-term memory (LSTM), and fully connected network (FCN) with channel attention mechanism. LSTM can capture temporal dependencies in the sequential data, conversely, CNN processes ECG signals and extracts spatial features. Feature selection is improved by introducing a Squeeze-andExcitation (SE) block that gives importance to more relevant channels and suppresses the irrelevant ones. This model is trained and further tested on seven types of ECG beats taken from the MIT-BIH arrhythmia dataset. Segmentation and noise filtering is used to improve the data quality. The proposed hybrid architecture achieves a 99.68% success rate and outperforms other deep learning and traditional methods. The results reveal that combining attention-based learning with spatial and temporal learning improves the classification performance and is suitable for use in healthcare applications that require real-time data.

Zitieren

APA

Kruthi, V. B., Kumari, L. R., & Rao, Y. C. (2026). Channel Attention based 1D-CNN-LSTM Fully Connected Network for ECG Beat Classification using SE Block. International Journal of Drug Delivery Technology, 16(34s). https://doi.org/10.25258/ijddt.16.34s.39

Vancouver

Kruthi VB, Kumari LR, Rao YC. Channel Attention based 1D-CNN-LSTM Fully Connected Network for ECG Beat Classification using SE Block. International Journal of Drug Delivery Technology. 2026;16(34s). doi:10.25258/ijddt.16.34s.39

BibTeX
@article{Kruthi2026,
  author = {Kruthi, V. Bhuvana and Kumari, L.V. Rajani and Rao, Y. Chalapathi},
  title = {Channel Attention based 1D-CNN-LSTM Fully Connected Network for ECG Beat Classification using SE Block},
  journal = {International Journal of Drug Delivery Technology},
  year = {2026},
  volume = {16},
  number = {34s},
  doi = {10.25258/ijddt.16.34s.39}
}
Kostenlos starten →