Predicting Age and Sex from Reduced Lead Electrocardiograms using Deep Learning

• Felipe M. Dias Hospital das Clínicas FMUSP / POLI USP
• Estela Ribeiro POLI USP
• Quenaz B. Soares POLI USP
• Jose E. Krieger POLI USP
• Marco A. Gutierrez Hospital das Clínicas FMUSP / POLI USP

DOI: https://doi.org/10.5753/sbcas.2025.7705

Resumo

Artificial intelligence is increasingly used to extract health insights from 12-lead (12L) electrocardiograms (ECG). Here, we propose a deeplearning model to predict sex and age from 12L and reduced-lead ECGs (6L–1L) and assess their impact on mortality risk. Using a ResNeXt-based model trained on the CODE15 dataset, our best models achieved an F1-score of 0.800 for sex classification (12L) and a mean absolute error of 8.961 for age estimation (4L). We found that overestimated age predictions and incorrect sex classifications were associated with higher mortality risk, whereas underestimated age predictions correlated with lower risk. These findings highlight the potential of reduced-lead ECGs for risk assessment, expanding their clinical utility.

Referências

AlGhatrif, M. and Lindsay, J. (2012). A brief review: history to understand fundamentals of electrocardiography. Journal of community hospital internal medicine perspectives, 2(1):14383.

Attia, Z. I., Friedman, P. A., Noseworthy, P. A., Lopez-Jimenez, F., Ladewig, D. J., Satam, G., Pellikka, P. A., Munger, T. M., Asirvatham, S. J., Scott, C. G., Carter, R. E., and Kapa, S. (2019a). Age and sex estimation using artificial intelligence from standard 12-lead ecgs. Circulation: Arrhythmia and Electrophysiology, 12(9):e007284.

Attia, Z. I., Kapa, S., Yao, X., Lopez-Jimenez, F., Mohan, T. L., Pellikka, P. A., Carter, R. E., Shah, N. D., Friedman, P. A., and Noseworthy, P. A. (2019b). Prospective validation of a deep learning electrocardiogram algorithm for the detection of left ventricular systolic dysfunction. Journal of cardiovascular electrophysiology, 30(5):668–674.

Batchvarov, V. N., Ghuran, A., Smetana, P., Hnatkova, K., Harries, M., Dilaveris, P., Camm, A. J., and Malik, M. (2002). Qt-rr relationship in healthy subjects exhibits substantial intersubject variability and high intrasubject stability. American Journal of Physiology-Heart and Circulatory Physiology, 282(6):H2356–H2363.

Cohen-Shelly, M., Attia, Z. I., Friedman, P. A., Ito, S., Essayagh, B. A., Ko, W.-Y., Murphree, D. H., Michelena, H. I., Enriquez-Sarano, M., Carter, R. E., et al. (2021). Electrocardiogram screening for aortic valve stenosis using artificial intelligence. European heart journal, 42(30):2885–2896.

Cox, D. R. (1972). Regression models and life-tables. Journal of the Royal Statistical Society: Series B (Methodological), 34(2):187–202.

He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778.

Lima, E. M., Ribeiro, A. H., Paixão, G. M. M., Ribeiro, M. H., Pinto-Filho, M. M., Gomes, P. R., Oliveira, D. M., Sabino, E. C., Duncan, B. B., Giatti, L., Barreto, S. M., Meira Jr, W., Schön, T. B., and Ribeiro, A. L. P. (2021). Deep neural network-estimated electrocardiographic age as a mortality predictor. Nat Commun, 12(1):5117.

Macfarlane, P., McLaughlin, S., Devine, B., and Yang, T. (1994). Effects of age, sex, and race on ecg interval measurements. Journal of Electrocardiology, 27:14–19. Research and Technology Transfer in Computerized Electrocardiology.

Macfarlane, P. W. and Kennedy, J. (2021). Automated ecg interpretation—a brief history from high expectations to deepest networks. Hearts, 2(4):433–448.

Malik, M., Hnatkova, K., Kowalski, D., Keirns, J. J., and van Gelderen, E. M. (2013). Qt/rr curvatures in healthy subjects: sex differences and covariates. American Journal of Physiology-Heart and Circulatory Physiology, 305(12):H1798–H1806.

PIPBERGER, H. V., FREIS, E. D., TABACK, L., and MASON, H. L. (1960). Preparation of electrocardiographic data for analysis by digital electronic computer. Circulation, 21(3):413–418.

Rafie, N., Kashou, A. H., and Noseworthy, P. A. (2021). Ecg interpretation: Clinical relevance, challenges, and advances. Hearts, 2(4):505–513.

Reyna, M. A., Sadr, N., Alday, E. A. P., Gu, A., Shah, A. J., Robichaux, C., Rad, A. B., Elola, A., Seyedi, S., Ansari, S., Ghanbari, H., Li, Q., Sharma, A., and Clifford, G. D. (2021). Will two do? varying dimensions in electrocardiography: The physionet/computing in cardiology challenge 2021. In 2021 Computing in Cardiology (CinC), volume 48, pages 1–4.

Ribeiro, A. H., Ribeiro, M. H., Paixão, G. M., Oliveira, D. M., Gomes, P. R., Canazart, J. A., Ferreira, M. P., Andersson, C. R., Macfarlane, P. W., Meira Jr, W., et al. (2020). Automatic diagnosis of the 12-lead ecg using a deep neural network. Nature communications, 11(1):1760.

Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., and Rabinovich, A. (2015). Going deeper with convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1–9.

Xie, S., Girshick, R., Dollár, P., Tu, Z., and He, K. (2017a). Aggregated residual transformations for deep neural networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1492–1500.

Xie, S., Girshick, R., Dollár, P., Tu, Z., and He, K. (2017b). Aggregated residual transformations for deep neural networks. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 5987–5995.