FEENet: A Frequency-Enhanced ECG Network for Cardiovascular Disease Detection
DOI:
https://doi.org/10.54097/hhavyk10Keywords:
Frequence-Enhance, Feature Pyramid, Transformer, Cardiovascular Disease, ElectrocardiographAbstract
With the development of society, the incidence of cardiovascular diseases has continued to rise. As an important non-invasive diagnostic tool, electrocardiography (ECG) has been widely used in clinical screening and auxiliary diagnosis. In recent years, deep learning-based automated ECG analysis methods have attracted considerable attention; however, most existing approaches rely on single time-domain features, making it difficult to fully exploit the discriminative information of ECG signals in both time and frequency domains. Therefore, this study proposes a Frequence-Enhance ECG Network (FEENet), which performs multi-scale feature extraction on both temporal ECG signals and their frequency-domain representations. A Transformer-based causal encoder (TCE) is then employed to model the relationships among features at different scales. Subsequently, a time–frequency cross-attention (TFCA) module is introduced to enable bidirectional interaction and deep fusion between temporal and frequency-domain features. Finally, a classification layer is used to produce accurate cardiovascular disease classification results. The proposed method achieves an accuracy of 89.6% and an F1-score of 77.4% on the PTB-XL dataset, demonstrating its effectiveness and strong classification capability in complex ECG classification tasks.
Downloads
References
[1] Alberdi A , Aztiria A , Basarab A .Towards an automatic early stress recognition system for office environments based on multimodal measurements: A review[J].Journal of Biomedical Informatics, 2016, 59:49-75.
[2] Yang S , Lian C , Zeng Z ,et al. A Multi-view Multi-scale Neural Network for Multi-label Ecg Classification [J].IEEE Transactions on Emerging Topics in Computational Intelligence, 2023, 7(3):648-660.
[3] Zhang H , Zhao W , Liu S .SE-ECGNet: A Multi-scale Deep Residual Network with Squeeze-and-Excitation Module for ECG Signal Classification[J]. 2020, 2685-2691.
[4] Srivastava A, Pratiher S, Alam S, et al. A deep residual inception network with channel attention modules for multi-label cardiac abnormality detection from reduced-lead ECG[J]. Physiological Measurement, 2022, 43(6): 064005.
[5] Wickramasinghe N L , Athif M .Multi-label classification of reduced-lead ECGs using an interpretable deep convolutional neural network[J].Physiological Measurement, 2022, 43(6):14.
[6] Wagner P , Strodthoff N , Bousseljot R D ,et al.PTB-XL, a large publicly available electrocardiography dataset [J].Scientific Data, 2020, 7(1):154.
[7] Hu J, Shen L, Sun G. Squeeze-and-Excitation Networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018: 7132-7141.
[8] Qin Y Z, Sun L, Chen H, et al. MVKT-ECG: Efficient single-lead ECG classification for multi-label arrhythmia by multi-view knowledge transferring[J]. Computers in Biology and Medicine, 2023, 166: 107503.
[9] Jyotishi D, Dandapat S. An attentive spatio-temporal learning-based network for cardiovascular disease diagnosis[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2023, 53(8): 4661-4671.
[10] Xiao Q, Li Y, Lee K, et al. Integrating learnable expert knowledge into deep learning-based multi-label ECG classification[J]. Measurement Science and Technology, 2025, 36(5): 056113.
[11] Yao Q H, Wang R X, Fan X M, et al. Multi-class arrhythmia detection from 12-lead varied-length ECG using attention-based time-incremental convolutional neural network[J]. Information Fusion, 2020, 53: 174-182.
[12] Jyotishi D, Dandapat S. An attention based hierarchical LSTM model for detection of myocardial infarction[C]//2020 IEEE 17th India Council International Conference (INDICON). New Delhi, India: IEEE, 2020: 1-5.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 International Journal of Advanced Engineering and Technology Research
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
