mHealth technology, by using habitual devices, i.e., smartphones, improves prevention, diagnosis, treatment, monitoring, and management of health. Monitoring heart profile during intense sports activities allows to diagnose pathologies, not identifiable with the traditional Holter approach and, therefore, it can help preventing possible injuries. On the other hand, denoising and extracting features from electrocardiographic (ECG) signal acquired during physical activity is a challenging task due to motion artifacts and measurement noise. In this paper, we propose a solution enabling a complete analysis of ECG signal through the implementation of a robust denoising algorithm, which has been characterized on synthetic signals and then has been tested on real traces acquired with a low-cost smartphone-based device during motorbike and car races.
Tobias Kober, Tom Hilbert, Jérôme Yerly, Ruud van Heeswijk, Elena Najdenovska
Denis Gillet, Man Shi, Jianwei Li