Complexity-Reduced End-to-End Fetal ECG Signal Recovery and QRS Complex Detection

• Julia C. Remus UFRGS
• Thiago L. T. da Silveira UFRGS / UFSM

DOI: https://doi.org/10.5753/sbcas.2025.7731

Resumo

Non-invasive electrocardiogram (niECG) enables pregnancy monitoring by assessing the mother’s and fetus’ health through maternal abdominal signal acquisition. However, isolating fetal ECG (fECG) is challenging due to low signal-to-noise ratio and time and frequency overlap with maternal cardiac activity. Prior studies focused on fetal R-peak detection and more recent works target full fECG waveform noise-free reconstruction, but at a high computational cost, precluding real-world applications. This paper explores different ways of reducing the complexity of state-of-the-art deep learning-based method for fECG recovery and fetal QRS complex localization. Results indicate that one encoder-decoder block pair can be removed without significantly impacting the metrics. While the other complexity-reduction options yield slightly lower metrics compared to the baseline and related works, they can be fine-tuned and adapted based on the specific requirements and objectives of the application.

Referências

Andreotti, F. et al. (2016). An open-source framework for stress-testing non-invasive foetal ECG extraction algorithms. Physiological Measurement, 37(5):627–648.

Barnova, K. et al. (2024). Artificial intelligence and machine learning in electronic fetal monitoring. Archives of Computational Methods in Engineering.

Behar, J. et al. (2016). A practical guide to non-invasive foetal electrocardiogram extraction and analysis. Physiological Measurement, 37(5):R1–R35.

Cohen, R. et al. (2024). Looks too good to be true: An information-theoretic analysis of hallucinations in generative restoration models. arXiv preprint arXiv:2405.16475.

Ghonchi, H. and Abolghasemi, V. (2022). A dual attention-based autoencoder model for fetal ECG extraction from abdominal signals. IEEE Sensors Journal, 22(23):22908–22918.

Jezewski, J. et al. (2012a). Determination of fetal heart rate from abdominal signals: evaluation of beat-to-beat accuracy in relation to the direct fetal electrocardiogram. Biomedizinische Technik Biomedical Engineering, 57(5).

Jezewski, J. et al. (2012b). Determination of fetal heart rate from abdominal signals: evaluation of beat-to-beat accuracy in relation to the direct fetal electrocardiogram. Biomedizinische Technik/Biomedical Engineering, 57(5).

Kahankova, R. et al. (2020). A review of signal processing techniques for non-invasive fetal electrocardiography. IEEE Reviews in Biomedical Engineering, 13:51–73.

Liu, Y. et al. (2025). Research on approximate computation of signal processing algorithms for aiot processors based on deep learning. Electronics, 14(6):1064.

Matonia, A. et al. (2020). Fetal electrocardiograms, direct and abdominal with reference heartbeat annotations. Scientific Data, 7(1).

Mendis, L. et al. (2023). Computerised cardiotocography analysis for the automated detection of fetal compromise during labour: A review. Bioengineering, 10(9):1007.

Mohebbian, M. R. et al. (2022). Fetal ECG extraction from maternal ECG using attention-based cyclegan. IEEE Journal of Biomedical and Health Informatics, 26(2):515–526.

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

Rahman, A. et al. (2023). Fetal ecg extraction from maternal ecg using deeply supervised LinkNet++ model. Engineering Applications of Artificial Intelligence, 123:106414.

Remus, J. C. and da Silveira, T. L. T. (2024). An end-to-end roi-based encoder-decoder for fetal ecg recovery and qrs complex detection. In International Symposium on Medical Measurements and Applications, volume 40, page 1–6. IEEE.

Sameni, R. (2021). Noninvasive fetal electrocardiography: Models, technologies, and algorithms. In Innovative Technologies and Signal Processing in Perinatal Medicine: Volume 1, volume 1, chapter 5. Springer International Publishing.

Shi, X. et al. (2023). Unsupervised learning-based non-invasive fetal ECG muti-level signal quality assessment. Bioengineering, 10(1):66.

Shuvo, M. M. H. et al. (2023). Efficient acceleration of deep learning inference on resource-constrained edge devices: A review. Proceedings of the IEEE, 111(1):42–91.

Silva, I. et al. (2013). Noninvasive fetal ECG: the physionet/computing in cardiology challenge 2013. Computing in cardiology, 40:149–152.

Sober, M. M. and Marco, J. G. (2007). Non-invasive fetal electrocardiogram database.

UNICEF (2019). Healthy mothers, healthy babies: Taking stock of maternal health. Technical report, United Nations Children’s Fund.

Vullings, R. et al. (2009). Dynamic segmentation and linear prediction for maternal ecg removal in antenatal abdominal recordings. Physiological Measurement, 30(3):291–307.

Wang, X. et al. (2023). Correlation-aware attention cyclegan for accurate fetal ecg extraction. IEEE Transactions on Instrumentation and Measurement, 72:1–13.

WHO, editor. Trends in maternal mortality 2000 to 2020. World Health Organization, Geneva.

Zhong, W. et al. (2019). Fetal electrocardiography extraction with residual convolutional encoder–decoder networks. Australasian Physical amp; Engineering Sciences in Medicine, 42(4):1081–1089.