ABSTRACT
The development of automated electrocardiogram
(ECG) analysis systems that are both accurate and clinically
interpretable is driven by the fact that cardiovascular diseases
continue to be a major cause of death worldwide. Although
many deep learning techniques function as black-box models
and provide little insight into the spatial reasoning across ECG
leads that supports clinical decision-making, recent methods
have demonstrated strong diagnostic performance on large-
scale ECG datasets. Moreover, Transformer-based architectures
introduce significant computational overhead when modelling
long-duration biosignals.
With an emphasis on graph-based learning and Selective State
Space Models (SSMs), this paper provides an organised overview
and methodological synthesis of recent developments in ECG
modelling. In addition to reviewing their extensions that include
convolutional front-ends, spectro-temporal embeddings, bidirec-
tional processing, and multi-branch tokenisation techniques, we
examine the rise of Mamba-based architectures as effective
substitutes for long-range temporal modelling. Simultaneously,
we investigate graph neural network formulations that explicitly
model inter-lead relationships, emphasising the growing interest
in dynamic graph learning and the drawbacks of static adjacency
assumptions.
We describe a Mamba-first, graph-second pipeline that sepa-
rates temporal feature extraction from spatial reasoning based
on these discoveries. In this framework, Mamba-based encoders
independently model lead-wise temporal dynamics, after which a
dynamically learned graph captures pathology-dependent inter-
lead interactions, yielding an interpretable reasoning matrix.
We demonstrate that hybrid spatio-temporal architectures can
achieve diagnostic performance comparable to strong convo-
lutional baselines while offering significantly improved trans-
parency through a quantitative comparison of representative
models and visual analysis of performance–interpretability trade-
offs.
All things considered, this work summarises recent develop-
ments in effective temporal modelling and comprehensible spatial
reasoning for ECG analysis and suggests promising avenues
for further investigation toward reliable, clinically useful ECG
decision-support systems.
Index Terms
Electrocardiography, State Space Models,
Mamba, Graph Neural Networks, Explainable AI, Arrhythmia
Classification.
REFERENCES
- [1] P. Wagner, N. Strodthoff, R.-D. Bousseljot, D. Kreiseler, F. I. Lunze, W. Samek, and T. Schaeffter, “Ptb-xl, a large publicly available electrocardiography dataset,” Scientific Data, vol. 7, no. 1, p. 154, 2020.
- [2] H. A. Elyamani, M. A. Salem, F. Melgani, and N. M. Yhiea, “Deep residual 2d convolutional neural network for cardiovascular disease classification,” Scientific Reports, vol. 14, p. 22040, 2024.
- [3] B. Andayeshgar and F. Abdali-Mohammadi, “Arrhythmia detection by the graph convolution network and a proposed structure for communication between cardiac leads,” BMC Medical Research Methodology, vol. 24, p. 96, 2024.
- [4] A. E. M. Atwa, E.-S. Atlam, A. Ahmed, M. A. Atwa, E. M. Abdelrahim, and A. I. Siam, “Interpretable deep learning models for arrhythmia classification based on ecg signals using ptb-xl dataset,” Diagnostics, vol. 15, no. 15, p. 1950, 2025.
- [5] Y. Chen, “A benchmark study of deep learning methods for multi-label pediatric electrocardiogram-based cardiovascular disease classification,” arXiv preprint arXiv:2510.03780, 2025.
- [6] A. Gu and T. Dao, “Mamba: Linear-time sequence modeling with selective state spaces,” arXiv preprint arXiv:2312.00752, 2023.
- [7] Z. Jiang et al., “One dimensional cnn ecg mamba for multilabel abnormality classification in 12 lead ecg,” arXiv preprint arXiv:2510.xxxxx, 2025.
- [8] H. Wang and T. Shen, “Visualized lead selection for arrhythmia classification based on a lead activation heatmap using multi-lead ecgs,” Bioengineering, vol. 11, no. 6, p. 578, 2024.
- [9] J. Hu, M. M. U. Rahman, and T.-M. Laleg-Kirati, “Adaptive long-range modeling of eeg and ecg with mamba and dynamic graph learning,” Scientific Reports, vol. 15, p. 38762, 2025.
- [10] Y. Li et al., “Dyg-mamba: Continuous state space modeling on dynamic graphs,” arXiv preprint arXiv:240x.xxxxx, 2024.
- [11] L. Li, H. Wang, and W. Zhang, “Stg-mamba: Spatial-temporal graph learning via selective state space model,” arXiv preprint arXiv:2403.12418, 2024.
- [12] X. Wang, J. Kang, P. Han, Y. Zhao, Q. Liu, L. He, L. Zhang, L. Dai, Y. Wang, and J. Tao, “Ecg-expert-qa: A benchmark for evaluating medical large language models in heart disease diagnosis,” IEEE Journal of Biomedical and Health Informatics, 2025.
- [13] A. Lindenmeyer, M. Blattmann, S. Franke, T. Neumuth, and D. Schneider, “Towards trustworthy ai in healthcare: Epistemic uncertainty estimation for clinical decision support,” Journal of Personalized Medicine, vol. 15, no. 2, p. 58, 2025.
- [14] A. Lukyanenko et al., “Deep survival analysis from adult and pediatric electrocardiograms: A multi-center benchmark study,” arXiv preprint arXiv:250x.xxxxx, 2025.
- [15] Z. Jiang et al., “Ecg-mamba: Cardiac abnormality classification with non-uniform-mix augmentation on 12-lead ecgs,” arXiv preprint arXiv:2509.xxxxx, 2025.
- [16] J. Q. T. L. Goh, B. Xie, C. Zhu, B. Wan, Y. Guan, R. D. Chen, and P. Y. Chiang, “Biomamba: Leveraging spectro-temporal embedding in bidirectional mamba for enhanced biosignal classification,” arXiv preprint arXiv:2503.11741, 2025.
- [17] L. Zhu, B. Liao, Q. Zhang, X. Wang, W. Liu, and X. Wang, “Vision mamba: Efficient visual representation learning with bidirectional state space model,” Proceedings of ICML, 2024.
- [18] T. Dao and A. Gu, “Transformers are ssms: Generalized models and efficient algorithms through structured state space duality,” arXiv preprint arXiv:2405.21060, 2024.
- [19] H. Zhang, R. Wang, C. Zhou, J. Shi, Y. Ge, Z. Li, S. Chang, H. Wang, J. He, and Q. Huang, “S2m2ecg: Spatio-temporal bi-directional state space model enabled multi-branch mamba for ecg,” arXiv preprint arXiv:2509.03066, 2025.
- [20] Y. Xu, X. Liu, Z. Kong, Y. Wu, Y. Wang, Y. Lu, and H. Xu, “Mambacapsule: Towards transparent cardiac disease diagnosis with electrocardiography using mamba capsule network,” arXiv preprint arXiv:2407.20893, 2024.
- [21] Z. Wang et al., “Graph-mamba: Towards long-range graph sequence modeling with selective state spaces,” arXiv preprint arXiv:2402.08678, 2024.
- [22] A. Behrouz and F. Hashemi, “Graph mamba: Towards learning on graphs with state space models,” arXiv preprint arXiv:2402.08678, 2024.
- [23] Y. Li et al., “Exploring graph mamba: A comprehensive survey on state space models for graph learning,” arXiv preprint arXiv:240x.xxxxx, 2024.
- [24] W. Li et al., “A reliable deep learning model for ecg interpretation: Mitigating overconfidence and direct uncertainty quantification,” arXiv preprint arXiv:250x.xxxxx, 2025.
- [25] Z. Huang, N. Sahoo, A. Kumari, G. Kumar, K. Cai, S. Cao, Y. Kang, T. Xia, S. Chatterjee, N. Hausman, A. Jay, E. S. Rosenthal, S. Srinivasan, S. Hasan, A. Fedorov, and S. Vesal, “High-fidelity synthetic ecg generation via mel-spectrogram informed diffusion training,” NeurIPS Workshop on GenAI for Health, 2025.