Quality Assessment of Photoplethysmography Signals For Cardiovascular Biomarkers Monitoring Using Wearable Devices

• Felipe M. Dias USP
• Marcelo A. F. Toledo USP
• Diego A. C. Cardenas USP
• Douglas A. Almeida USP
• Filipe A. C. Oliveira USP
• Estela Ribeiro USP
• Jose E. Krieger USP
• Marco Antonio Gutierrez USP

DOI: https://doi.org/10.5753/eniac.2025.11804

Abstract

Photoplethysmography (PPG) is a non-invasive technique widely used to monitor cardiovascular parameters such as heart rate. However, its reliability can be compromised by factors like motion artifacts. In this study, we extracted 27 statistical features from PPG signals and trained multiple machine learning models (XGBoost, CatBoost, Random Forest) to assess signal quality. Using a publicly available dataset of PPG time series, we evaluated model performance using sensitivity, positive predictive value, and F1-score. Our best model (CatBoost) achieved 94.7%, 95.9%, and 95.3% for these metrics, respectively. These results are comparable to state-of-the-art approaches but relying on relatively simple models.

References

Alian, A. A. and Shelley, K. H. (2014). Photoplethysmography. Best Practice & Research Clinical Anaesthesiology, 28(4):395–406.

Bishop, S. M. and Ercole, A. (2018). Multi-scale peak and trough detection optimised for periodic and quasi-periodic neuroscience data. In Heldt, T., editor, Intracranial Pressure & Neuromonitoring XVI, pages 189–195, Cham. Springer International Publishing.

Breiman, L. (2001). Random forests. Machine Learning, 45:5–32.

Charlton, P. H., Kyriacou, P. A., Mant, J., Marozas, V., Chowienczyk, P., and Alastruey, J. (2022). Wearable photoplethysmography for cardiovascular monitoring. Proceedings of the IEEE, 110(3):355–381.

Chen, T. and Guestrin, C. (2016). XGBoost. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM.

Dias, F. M., Costa, T. B., Cardenas, D. A. C., Toledo, M. A. F., Kriger, J. E., and Gutierrez, M. A. (2022). A machine learning approach to predict arterial blood pressure from photoplethysmography signal. In 2022 Computing in Cardiology (CinC), volume XX, pages 1–4.

Elgendi, M. (2012). On the analysis of fingertip photoplethysmogram signals. Current cardiology reviews, 8(1):14–25.

Elgendi, M. (2016). Optimal signal quality index for photoplethysmogram signals. Bioengineering, 3(4).

Fine, J., Branan, K. L., Rodriguez, A. J., Boonya-ananta, T., Ajmal, Ramella-Roman, J. C., McShane, M. J., and Coté, G. L. (2021). Sources of inaccuracy in photoplethys-mography for continuous cardiovascular monitoring. Biosensors, 11(4).

Goldberger, A. L., Amaral, L. A. N., Glass, L., Hausdorff, J. M., Ivanov, P. C., Mark, R. G., Mietus, J. E., Moody, G. B., Peng, C.-K., and Stanley, H. E. (2000). Physiobank, physiotoolkit, and physionet. Circulation, 101(23):e215–e220.

Li, S., Liu, L., Wu, J., Tang, B., and Li, D. (2018). Comparison and noise suppression of the transmitted and reflected photoplethysmography signals. BioMed research international, 2018.

Mejía-Mejía, E., Allen, J., Budidha, K., El-Hajj, C., Kyriacou, P. A., and Charlton, P. H. (2022). 4 - photoplethysmography signal processing and synthesis. In Allen, J. and Kyriacou, P., editors, Photoplethysmography, pages 69–146. Academic Press.

Mohagheghian, F., Han, D., Peitzsch, A., Nishita, N., Ding, E., Dickson, E. L., DiMezza, D., Otabil, E. M., Noorishirazi, K., Scott, J., Lessard, D., Wang, Z., Whitcomb, C., Tran, K.-V., Fitzgibbons, T. P., McManus, D. D., and Chon, K. H. (2022). Optimized signal quality assessment for photoplethysmogram signals using feature selection. IEEE Transactions on Biomedical Engineering, 69(9):2982–2993.

Moscato, S., Palmerini, L., Palumbo, P., and Chiari, L. (2022). Quality assessment and morphological analysis of photoplethysmography in daily life. Front. Digit. Health, 4:912353.

Mukkamala, R., Hahn, J.-O., and Chandrasekhar, A. (2022). 11 - photoplethysmography in noninvasive blood pressure monitoring. In Allen, J. and Kyriacou, P., editors, Photoplethysmography, pages 359–400. Academic Press.

Nitzan, M. and Ovadia-Blechman, Z. (2022). 9 - physical and physiological interpretations of the ppg signal. In Allen, J. and Kyriacou, P., editors, Photoplethysmography, pages 319–340. Academic Press.

Prokhorenkova, L., Gusev, G., Vorobev, A., Dorogush, A. V., and Gulin, A. (2017). Catboost: unbiased boosting with categorical features.

Reisner, A., Shaltis, P., McCombie, D., Asada, H., Warner, D., and Warner, M. (2008). Utility of the Photoplethysmogram in Circulatory Monitoring. Anesthesiology, 108(5):950–958.

Torres-Soto, J. and Ashley, E. A. (2020). Multi-task deep learning for cardiac rhythm detection in wearable devices. npj Digital Medicine, 3(1):1–8.