Heart failure classification in echocardiograms using Deep Neural Networks

• Luiz Fernando Oliveira Maciel PUC Minas
• Alexei Manso Correa Machado PUC Minas / UFMG

DOI: https://doi.org/10.5753/sbcas_estendido.2025.6965

Abstract

This article presents a deep learning model based on MobileNetv4 for detecting heart failure in echocardiograms based on left ventricle ejection fraction (LVEF) estimation. The pipeline consists of two stages: a regression model predicting ventricular volume coordinates, and a classification model that computes LVEF from the predicted volumes. A 5-fold cross-validation was performed, yielding a coefficient of determination R² of 0.83 ± 0.01 (95% CI: [0.83, 0.84]) and an accuracy of 0.74 ± 0.01 (95% CI: [0.74, 0.75]).

References

Ghorbani, A., Ouyang, D., Abid, A., He, B., Chen, J., Harrington, R., Liang, D., Ashley, E., and Zou, J. (2020). Deep learning interpretation of echocardiograms. NPJ digital medicine, 3(1):10.

McGill, H., McMahan, A., and Gidding, S. (2008). Preventing heart disease in the 21st century. Circulation, 117(9):1216–27.

Omerovic, S. and Jain, A. (2024). Echocardiogram. In StatPearls. StatPearls Publishing, Treasure Island (FL).

OMS (2021). Global health estimates: Leading causes of death.

O’Shea, K. and Nash, R. (2015). An introduction to convolutional neural networks. arXiv.

Ouyang, D., He, B., Ghorbani, A., Yuan, N., Ebinger, J., Langlotz, C., Heidenreich, P., Harrington, R., Liang, D., Ashley, E., et al. (2020). Video-based ai for beat-to-beat assessment of cardiac function. Nature, 580(7802):252–6.

Paul, A. and Bhuiyan, Y. (2024). EchoTrace: A 2D echocardiography deep learning approach for left ventricular ejection fraction prediction. J. Electron. Electric. Eng.

Qin, D., Leichner, C., Delakis, M., Fornoni, M., Luo, S., Yang, F., Wang, W., Banbury, C., Ye, C., Akin, B., et al. (2024). Mobilenetv4-universal models for the mobile ecosystem. arXiv preprint arXiv:2404.10518.