05 / 2023 – 10 / 2023
The application of wearables in medicine has increased exponentially in the last few years due to the growing number of features that can monitor health parameters in real time  . In everyday life settings, consumers use smartwatches to monitor their heart rate or heart rate variability , whereas medical staff could benefit from features such as the ability to record a single-lead electrocardiogram (ECG). Being able to record an ECG similar to lead I on a standard 12-lead ECG on a smartwatch opens up a wide range of possibilities in remote assessment and monitoring of patients , especially those with heart diseases such as heart failure (HF). Thus, continuous cardiac monitoring at home could potentially enable proactive care , and the detection of heart malfunctions before serious health problems occur .
The data from ECG tracings made by a smartwatch is not as extensive as the data from a 12-lead ECG . Additionally, the algorithms supplied with these wearables show insufficient performance in segmenting the ECG signals compared to manual 12-lead measurements. In particular, the Withings Scanwatch (WS) and the Apple Watch (AW) currently only screen for atrial fibrillation (AF)  .
For further analysis, it is important to be able to detect abnormalities in ECG characteristics such as arrhythmia, negative P, T, and R peaks , ST elevation, or non-ST elevation, as they might be signs of acute coronary syndromes . Studies showed that the current segmentation algorithm of the WS failed to detect the QT interval in 44% of measurements  and the AW has an insufficient identification of P-waves . Thus, this thesis aims to implement, evaluate and compare different open-source ECG segmentation algorithms on their ability to segment ECGs measured by two different smartwatches, the WS and the AW. Consequently, the research question of this thesis is how successfully existing open-source algorithms can segment ECG tracings recorded by smartwatches. In light of that objective, the proposed work consists of the following parts:
• Literature and patent research of relevant work resulting in a comprehensive list of existing ECG segmentation algorithms and algorithm benchmarking
• Development of a suitable performance evaluation and benchmarking strategy for the segmentation of smartwatch ECGs
• Implementation of at least three open-source ECG segmentation algorithms
• Ground truth labeling of existing smartwatch ECG data set from HF patients and healthy individuals
• Evaluation of selected algorithms, also focusing on performance differences between healthy individuals and HF patients
 M. Strik, S. Ploux, D. Weigel, J. van der Zande, A. Velraeds, H.-P. Racine, F. D. Ramirez, M. Haïssaguerre, and P. Bordachar, “The use of smartwatch electrocardiogram beyond arrhythmia detection,” Trends in Cardiovascular Medicine, 2023. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S1050173822001530
 C. Ford, C. X. Xie, A. Low, K. Rajakariar, A. N. Koshy, J. K. Sajeev, L. Roberts, B. Pathik, and A. W. Teh, “Comparison of 2 smart watch algorithms for detection of atrial fibrillation and the benefit of clinician interpretation: SMART WARS study,” JACC: Clinical Electrophysiology, vol. 8, no. 6, pp. 782–791, 2022. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S2405500X22002201
 F. Sarhaddi, K. Kazemi, I. Azimi, R. Cao, H. Niela-Vilén, A. Axelin, P. Liljeberg, and A. M. Rahmani, “A comprehensive accuracy assessment of samsung smartwatch heart rate and heart rate variability,” PLOS ONE, vol. 17, no. 12, p. e0268361, 2022, publisher: Public Library of Science. [Online]. Available: https://journals.plos.org/plosone/article?id=10.1371/ journal.pone.0268361
 A. L. Bui and G. C. Fonarow, “Home monitoring for heart failure management,” vol. 59, no. 2, pp. 97–104, 2021, publisher: American College of Cardiology Foundation. [Online]. Available: https://www.jacc.org/doi/full/10.1016/j.jacc.2011.09.044
 D. Mannhart, E. Hennings, M. Lischer, C. Vernier, J. Du Fay de Lavallaz, S. Knecht, B. Schaer, S. Osswald, M. Kühne, C. Sticherling, and P. Badertscher, “Clinical validation of automated corrected QT-interval measurements from a single lead electrocardiogram using a novel smartwatch,” Frontiers in Cardiovascular Medicine, vol. 9, p. 906079, 2022. [Online]. Available: https://www.frontiersin.org/articles/10.3389/fcvm.2022.906079
 M. Hammad, A. Maher, K. Wang, F. Jiang, and M. Amrani, “Detection of abnormal heart conditions based on characteristics of ECG signals,” vol. 125, pp. 634–644. [Online]. Available: https://linkinghub.elsevier.com/retrieve/pii/S0263224118304147
 C. A. M. Spaccarotella, A. Polimeni, S. Migliarino, E. Principe, A. Curcio, A. Mongiardo, S. Sorrentino, S. De Rosa, and C. Indolfi, “Multichannel Electrocardiograms Obtained by a Smartwatch for the Diagnosis of ST-Segment Changes,” JAMA Cardiology, vol. 5, no. 10, pp. 1176–1180, 10 2020. [Online]. Available: https://doi.org/10.1001/jamacardio.2020.3994