Deep Learning Based Beat-to-Beat Delineation of Heart Sounds and Fiducial Points in Seismocardiography

Emil Korsgaard^1*Ahmad Agam²Peter Søgaard²Kasper Janus Grønn Emerek²Kasper Sørensen³Jørn Wulff Helge⁴Johannes J. Struijk¹Samuel Schmidt¹

• ¹Aalborg Universitet, Aalborg, Denmark
• ²Aalborg Universitetshospital, Aalborg, Denmark
• ³Ventriject Aps, Hellerup, Denmark
• ⁴Kobenhavns Universitet, Copenhagen, Denmark

Introduction: The application of deep learning methods in automatic delineation of fiducial points in the seismocardiography (SCG) on a beat-to-beat basis provides the possibility of obtaining a novel and comprehensive approach to assess and monitor myocardial mechanics and hemodynamic status. Therefore, the aim of this study is to develop an adaptive and data-driven algorithm for automatic delineation of 11 fiducial points in the SCG. Methods: SCG signals from both subjects with and without known cardiac disease (CD) were included. A semi-automatic annotation-pipeline was prepared for effective annotation of fiducial points for each individual cardiac cycle, where 42,452 individual beats from 198 subjects were annotated. A U-Net architecture deep learning model was developed to detect 11 fiducial points and predict multiple time intervals in the SCG signal. The evaluation metrics were positive predictive value and sensitivity. Results: The median positive predictive value and sensitivity of the algorithm ranged between 0.809 and 1.000 and 0.843 and 0.918 for different fiducial points, respectively. Conclusion: A novel algorithm for automatic detection of 11 fiducial points in the SCG was developed and tested in subjects both with and without CD.