Classificação de Doença Arterial Coronariana através de redes neurais profundas

• Diogo Casal Ferreira PUC Minas
• Alexei Manso Correa Machado PUC Minas / UFMG

DOI: https://doi.org/10.5753/sbcas.2025.6951

Resumo

As doenças arteriais coronarianas (DAC) figuram entre as principais causas de mortalidade no mundo, demandando avanços no diagnóstico precoce. Este estudo compara redes neurais LSTM, Bi-GRU e Transformer na classificação de eletrocardiogramas (ECG) em três categorias: normal, infarto inferior (IMI) e anterior (AMI), utilizando o conjunto de dados PTB-XL. Os resultados indicam que Transformers são mais rápidos no treinamento (média de 35 minutos), enquanto LSTM e Bi-GRU apresentam maior precisão (91% e 92%, respectivamente). O trabalho também contribui para a melhoria do préprocessamento e do treinamento de dados de ECG aplicados ao diagnóstico cardíaco.

Referências

M. A. Atiea e M. Adel. Transformer-based neural network for electrocardiogram classification. International Journal of Advanced Computer Science and Applications, 13 (11), 2022.

F. S. Butt, M. F. Wagner, J. Schäfer, e D. G. Ullate. Toward automated feature extraction for deep learning classification of electrocardiogram signals. IEEE Access, 10:118601–118616, 2022.

E. F. Carneiro. O eletrocardiograma: 10 anos depois. In O Eletrocardiograma: 10 anos depois, pages 622–622. 1997.

J. Chung. Empirical evaluation of gated recurrent neural networks on sequence modeling. arXiv preprint arXiv:1412.3555, 2014.

A. Darmawahyuni, S. Nurmaini, e Sukemi. Deep learning with long short-term memory for enhancement myocardial infarction classification. In 2019 6th International Conference on Instrumentation, Control, and Automation (ICA), pages 19–23, 2019. DOI: 10.1109/ICA.2019.8916683.

M. Dey, N. Omar, e M. A. Ullah. Temporal feature-based classification into myocardial infarction and other cvds merging cnn and bi-lstm from ecg signal. IEEE Sensors Journal, 21(19):21688–21695, 2021. DOI: 10.1109/JSEN.2021.3079241.

Q. Geng, H. Liu, T. Gao, R. Liu, C. Chen, Q. Zhu, e M. Shu. An ecg classification method based on multi-task learning and cot attention mechanism. Healthcare, 11(7), 2023. ISSN 2227-9032. DOI: 10.3390/healthcare11071000. URL [link].

S. Hochreiter. Long short-term memory. Neural Computation MIT-Press, 1997. E. B. Komilovich. Coronary artery disease. EUROPEAN JOURNAL OF MODERN MEDICINE AND PRACTICE, 3(12):81–87, 2023.

A. Natarajan, Y. Chang, S. Mariani, A. Rahman, G. Boverman, S. Vij, e J. Rubin. A wide and deep transformer neural network for 12-lead ecg classification. In 2020 Computing in Cardiology, pages 1–4. IEEE, 2020.

J. Pan e W. J. Tompkins. A real-time qrs detection algorithm. IEEE Transactions on Biomedical Engineering, BME-32(3):230–236, 1985. DOI: 10.1109/TBME.1985.325532.

N. Strodthoff, T. Mehari, C. Nagel, P. J. Aston, A. Sundar, C. Graff, J. K. Kanters, W. Haverkamp, O. Dössel, A. Loewe, et al. Ptb-xl+, a comprehensive electrocardiographic feature dataset. Scientific data, 10(1):279, 2023.

A. Vaswani. Attention is all you need. arXiv preprint arXiv:1706.03762, 2017. P. Wagner, N. Strodthoff, R.-D. Bousseljot, D. Kreiseler, F. I. Lunze, W. Samek, e T. Schaeffter. Ptb-xl, a large publicly available electrocardiography dataset. Scientific data, 7 (1):1–15, 2020.

P. Xiong, S. M.-Y. Lee, e G. Chan. Deep learning for detecting and locating myocardial infarction by electrocardiogram: A literature review. Frontiers in cardiovascular medicine, 9:860032, 2022.

X. Zhang, R. Li, H. Dai, Y. Liu, B. Zhou, e Z. Wang. Localization of myocardial infarction with multi-lead bidirectional gated recurrent unit neural network. IEEE Access, 7: 161152–161166, 2019a. DOI: 10.1109/ACCESS.2019.2946932.

X. Zhang, R. Li, Q. Hu, B. Zhou, e Z. Wang. A new automatic approach to distinguish myocardial infarction based on lstm. In 2019 8th International Symposium on Next Generation Electronics (ISNE), pages 1–3, 2019b. DOI: 10.1109/ISNE.2019.8896550.