Biological Data

Hybrid CNN-Transformer Models for ECG-Based Cardiac Diagnosis

Document Type : Review Article

Authors

Department of Computer Engineering, University of Mohaghegh Ardabili

Abstract

This review evaluates hybrid Convolutional Neural Network (CNN) and Transformer architectures applied to ECG-based cardiac monitoring. We analyze attention-based and multimodal frameworks using reported metrics from benchmark datasets, including MIT-BIH and PhysioNet. Because current research often restricts its scope to isolated architectures or single diagnostic tasks, this paper categorizes established methods for model selection, data preprocessing, and performance evaluation in single-lead and multi-lead ECG analysis. Published results show that hybrid configurations achieve peak accuracies of 98.6%, exceeding the ~95.0% typical of standalone CNNs and the ~96.0% of standalone Transformers. We conclude by examining limitations in data volume and cross-cohort generalization, alongside the requirements for multi-center clinical deployment.

Keywords

Abdollahi, J. (2026). HHO-optimized VAE-transformer framework for robust clinical phenotyping and EHR reconciliation in diabetes management. Discover Artificial Intelligence, 6, 1717. https://doi.org/10.1007/s44163-026-01815-x
Aburass, S., Dorgham, O., Al Shaqsi, J., Abu Rumman, M., & Al-Kadi, O. (2025). Vision transformers in medical imaging: a comprehensive review of advancements and applications across multiple diseases. Journal of Imaging Informatics in Medicine, 38(6), 3928-3971. https://doi.org/10.1007/s10278-025-01481-y
Agatra, D.F. N., Cornella, B.A., Muza'in, M., Munsarif, M., Ilham, A., & Abdollahi, J.  (2024). Enhancing Agricultural Pest Detection with EfficientNetV2-L and Grad-CAM: A Comprehensive Approach to Sustainable Farming. Journal of Intelligent Computing & Health Informatics, 5(1), 25-34. https://doi.org/10.26714/jichi.v5i1.13959
Akinlade, O., Abdollahi, J., Paul Etchie, M.P., Ajagbe, S.A., Alabi, O.O., & Ayo Adeyanju, B.A. (2026). Ensuring effective cervical cancer diseases diagnosis system using ensemble machine learning model. IAES International Journal of Artificial Intelligence, 15(3), 2410-2422. https://doi.org/10.11591/ijai.v15.i3.pp2410-2422 
Alkhodari, M., Hadjileontiadis, L.J., & Khandoker, A.H. (2024). Identification of congenital valvular murmurs in young patients using deep learning-based attention transformers and phonocardiograms. IEEE Journal of Biomedical and Health Informatics, 28(4), 1803-1814. https://doi.org/10.1109/JBHI.2024.3357506
Avola, D., Cannistraci, I., Cascio, M., Cinque, L., Fagioli, A., Foresti, G.L., Rodolà, E., & Solito, L. (2024). MV-MS-FETE: Multi-view multi-scale feature extractor and transformer encoder for stenosis recognition in echocardiograms. Computer Methods and Programs in Biomedicine, 245, 108037. https://doi.org/10.1016/j.cmpb.2024.108037
Ayano, Y.M., Schwenker, F., Dufera, B.D., Debelee, T.G., & Ejegu, Y.G. (2024). Interpretable hybrid multichannel deep learning model for heart disease classification using 12-lead ECG signal. IEEE Access, 12, 94055-94080. https://doi.org/10.1109/ACCESS.2024.3421641
Bai, X., Dong, X., Li, Y., Liu, R., & Zhang, H. (2024). A hybrid deep learning network for automatic diagnosis of cardiac arrhythmia based on 12-lead ECG. Scientific Reports, 14(1), 24441. https://doi.org/10.1038/s41598-024-75531-w
Busia, P., Scrugli, M.A., Jung, V. J.B., Benini, L., & Meloni, P. (2024). A tiny transformer for low-power arrhythmia classification on microcontrollers. IEEE Transactions on Biomedical Circuits and Systems, 19(1), 142-152. https://doi.org/10.1109/TBCAS.2024.3401858
Che, C., Zhang, P., Zhu, M., Qu, Y., & Jin, B. (2021). Constrained transformer network for ECG signal processing and arrhythmia classification. BMC Medical Informatics and Decision Making, 21, 184. https://doi.org/10.1186/s12911-021-01546-2
Chen, J., Huang, S., Zhang, Y., Chang, Q., Zhang, Y., Li, D., Qiu, J., Hu, L., Peng, X., Du, Y., Gao, Y., Chen, D.Z., Bellou, A., Wu, J., & Liang, H. (2024). Congenital heart disease detection by pediatric electrocardiogram based deep learning integrated with human concepts. Nature Communications, 15, 976. https://doi.org/10.1038/s41467-024-44930-y 
Chen, W., Luo, Y., & Wang, J. (2024). Three-branch temporal-spatial convolutional transformer for motor imagery EEG classification. IEEE Access, 12, 79754-79764. https://doi.org/10.1109/ACCESS.2024.3405652
Din, S., Qaraqe, M., Mourad, O., Qaraqe, K., & Serpedin, E. (2024). ECG-based cardiac arrhythmias detection through ensemble learning and fusion of deep spatial–temporal and long-range dependency features. Artificial Intelligence in Medicine, 150, 102818. https://doi.org/10.1016/j.artmed.2024.102818
Fredj, H.B., Fadhel, Y.B., & Souani, C. (2025). Medical images classification using convolutional neural network. In 2025 IEEE 22nd International Multi-Conference on Systems, Signals & Devices (SSD), https://doi.org/10.1109/SSD64182.2025.10990011
Guo, M.H., Lu, C.Z., Hou, Q., Liu, Z., Cheng, M.M., & Hu, S.M. (2022). Segnext: Rethinking convolutional attention design for semantic segmentation. Advances in Neural Information Processing Systems, 35, 1140-1156.
https://doi.org/10.48550/arXiv.2209.08575
Houssein, E.H., Mohamed, R.E., Hu, G., & Ali, A.A. (2024). Adapting transformer-based language models for heart disease detection and risk factors extraction. Journal of Big Data, 11, 47. https://doi.org/10.1186/s40537-024-00903-y 
Kim, J.W., Khan, A.U., & Banerjee, I. (2025). Systematic review of hybrid vision transformer architectures for radiological image analysis. Journal of Imaging Informatics in Medicine, 38(5), 3248-3262. https://doi.org/10.1007/s10278-024-01322-4 
Lawhern, V.J., Solon, A.J., Waytowich, N.R., Gordon, S.M., Hung, C.P., & Lance, B.J. (2018). EEGNet: a compact convolutional neural network for EEG-based brain–computer interfaces. Journal of Neural Engineering, 15(5), 056013. https://doi.org/10.1088/1741-2552/aace8c 
Li, B., Yang, B., Guan, C., & Hu, C. (2019). Three-class motor imagery classification based on fbcsp combined with voting mechanism. In 2019 IEEE International Conference on Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA), https://doi.org/10.1109/CIVEMSA45640.2019.9071618 
Li, J., & Zhang, Y. (2024). Regressive vision transformer for dog cardiomegaly assessment. Scientific Reports, 14, 1539. https://doi.org/10.1038/s41598-023-50063-x
Li, S., Wu, C., & Xiong, N. (2022). Hybrid architecture based on CNN and transformer for strip steel surface defect classification. Electronics, 11(8), 1200. https://doi.org/10.3390/electronics11081200 
Liu, X., He, L., Yan, J., Huang, Y., Wang, Y., Lin, C., Huang, Y., & Liu, X. (2024). A neural network for high-precise and well-interpretable electrocardiogram classification. Biorxiv, 1-9. https://doi.org/10.1101/2024.01.03.573822 
Mehrpour, O., Nakhaee, S., Abdollahi, J., & Vohra, V. (2025). Predictive modeling of methadone poisoning outcomes in children≤ 5 years: utilizing machine learning and the National Poison Data System for improved clinical decision-making. European Journal of Pediatrics, 184(2), 186. https://doi.org/10.1007/s00431-024-05957-x
Peng, X., Shu, W., Pan, C., Ke, Z., Zhu, H., Zhou, X., & Song, W.W. (2022). DSCSSA: A classification framework for spatiotemporal features extraction of arrhythmia based on the Seq2Seq model with attention mechanism. IEEE Transactions on Instrumentation and Measurement, 71, 1-12. https://doi.org/10.1109/TIM.2022.3194906
Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. Medical Image Computing and Computer-assisted Intervention – MICCAI 2015, 234-241. https://doi.org/10.1007/978-3-319-24574-4-28 
Sakhavi, S., Guan, C., & Yan, S. (2018). Learning temporal information for brain-computer interface using convolutional neural networks. IEEE Transactions on Neural Networks and Learning Systems, 29(11), 5619-5629. https://doi.org/10.1109/TNNLS.2018.2789927
Shashikumar, S.P., Shah, A.J., Clifford, G.D., & Nemati, S. (2018). Detection of paroxysmal atrial fibrillation using attention-based bidirectional recurrent neural networks. Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining,
https://doi.org/10.48550/arXiv.1805.09133
Wang, C., Ma, J., Wei, G., & Sun, X. (2025). Analysis of cardiac arrhythmias based on ResNet-ICBAM-2DCNN dual-channel feature fusion. Sensors, 25(3), 661. https://doi.org/10.3390/s25030661
Wang, J., Li, J., Wang, R., & Zhou, X. (2024). VAE-driven multimodal fusion for early cardiac disease detection. IEEE Access, 12, 90535-90551. https://doi.org/10.1109/ACCESS.2024.3420444
Wang, Z., Stavrakis, S., & Yao, B. (2023). Hierarchical deep learning with generative adversarial network for automatic cardiac diagnosis from ECG signals. Computers in Biology and Medicine, 155, 106641. https://doi.org/10.1016/j.compbiomed.2023.106641
Wieneke, H., & Voigt, I. (2024). Principles of artificial intelligence and its application in cardiovascular medicine. Clinical Cardiology, 47(1), e24148. https://doi.org/10.1002/clc.24148
Wu, H., Qu, G., Xiao, Z., & Chunyu, F. (2024). Enhancing left ventricular segmentation in echocardiography with a modified mixed attention mechanism in SegFormer architecture. Heliyon, 10(15), e34845. https://doi.org/10.1016/j.heliyon.2024.e34845
Xu, Y., Kong, Z., Wu, Y., Wang, Y., Liu, X., Lu, Y., Wu, J., & Xu, H. (2026). MambaCapsule: Toward Transparent cardiac disease diagnosis with electrocardiography using mamba capsule network. IEEE Transactions on Computational Social Systems, 14(8), 1-15.
https://doi.org/10.48550/arXiv.2407.20893
Zhang, C., Kim, Y.K., & Eskandarian, A. (2021). EEG-inception: an accurate and robust end-to-end neural network for EEG-based motor imagery classification. Journal of Neural Engineering, 18(4), 046014. https://doi.org/10.1088/1741-2552/abed81 
Zhang, J., Liu, A., Gao, M., Chen, X., Zhang, X., & Chen, X. (2020). ECG-based multi-class arrhythmia detection using spatio-temporal attention-based convolutional recurrent neural network. Artificial Intelligence in Medicine, 106, 101856.  https://doi.org/10.1016/j.artmed.2020.101856
Zhang, Z., Govindarajan, V., Ouyang, X., Yang, J., Alabdullah, B., Alammar, S., Por, L. Y., Guo, Q., & Li, M. (2026). CardioGPT-ViT: Explainable deep learning models with multimodal data for coronary artery disease and heart failure on edge consumer devices. IEEE Transactions on Consumer Electronics, 72(3), 6951-6963. https://doi.org/10.1109/TCE.2026.3667507
Zhou, Z., Rahman Siddiquee, M.M., Tajbakhsh, N., & Liang, J. (2018). Unet++: A nested u-net architecture for medical image segmentation. In Deep learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support, 3-11. https://doi.org/10.1007/978-3-030-00889-5-1 
Zhu, J., Feng, Y., Liu, Q., Xu, H., Miao, Y., Lin, Z., Li, J., Liu, H., Xu, Y., & Li, F. (2024). An improved convnext with multimodal transformer for physiological signal classification. IEEE Access, 12, 11217-11229. https://doi.org/10.1109/ACCESS.2024.3355273
Biological Data
Volume 1, Issue 1 - Serial Number 1
June 2026
Pages 104-135
  • Receive Date: 02 May 2026
  • Revise Date: 30 May 2026
  • Accept Date: 18 June 2026
  • First Publish Date: 22 June 2026