Copyright © 2025 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{170692,
        author = {Dipanjan Dutta and Anusudha.K},
        title = {A Systematic Review of Deep Learning Techniques for ECG Arrhythmia Classification},
        journal = {International Journal of Innovative Research in Technology},
        year = {2024},
        volume = {11},
        number = {7},
        pages = {515-524},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=170692},
        abstract = {The use of Deep Learning (DL) methods for detecting arrhythmias from ECG data, focusing on their clinical potential while highlighting areas that need more research for reliable application. The study stresses the importance of using diverse ECG datasets, advanced noise reduction, data augmentation, and new integrated DL models for better accuracy in clinical settings. Specifically, the authors introduce a hybrid model that combines Convolutional Neural Networks (CNN) with Variational Autoencoders (VAE) — the first use of VAE for classifying 1-D ECG signals. This model uses both supervised and unsupervised learning and shows better performance compared to standard CNNs and Autoencoders (AEs), particularly in reducing errors in diagnosing different cardiovascular diseases (CVDs). The paper also addresses challenges like data imbalance, which can be improved by using real-time data and augmentation techniques, thereby increasing model accuracy. Furthermore, In some paper we come across  developing a mobile app based on this CNN-VAE model, which could help healthcare professionals predict high-risk CVD patients. The review also covers successful DL algorithms, such as GRU, LSTM, DBM, BiLSTM and CNN, which have proven effective in classifying arrhythmias like atrial fibrillation and ventricular ectopic beats. The study highlights the importance of selecting the right DL algorithm to achieve optimal classification results.},
        keywords = {ECG, diagnostics, clinical decision support, CNNs, DBM, hybrid DL, RNNs, including LSTM and BiLSTM, GRU.},
        month = {December},
        }