Seismocardiography On Infants And Kids

Nico Jähne-Raden1, Henrike Gütschleg1, Marie Cathrine Wolf1, Stephan Sigg2, Ulf Kulau3
1Peter L. Reichertz Institute for Medical Informatics University of Braunschweig - Institute of Technology and Hannover Medical School, 2Department of Communications and Networking, 3DSI Aerospace


Abstract

Seismocardiography (SCG) offers a variety of possibilities for cardiovascular diagnostics, especially in long-term monitoring. Unfortunately, little infor-mation is available about changes in SCG-signal-morphology over the life course, as well as differences in different age groups. Physiologically, the change in the human body and notably in the heart has been extensively investigated with increasing age (e.g. doubled oxygen demand, heart size ratios differs). From birth to adolescence, the resting heart rate drops (about twice as high in new-borns) and the ejection volume increases as the body grows. This work is intended to provide a descriptive representation of changes in the SCG signal using an optimized measurement system. A highly specialized FPGA-based system with two isochronous accelerometers (Kio-nix-KX-122, at sternum/apex) and a reference ECG with a sample frequency of 17kHz was used. We included healthy infants and children between 0 and 14 years, with a focus on babies (n = 12, w = 6, m = 6, 4 children under one year). In contrast to adult measurements, working with children is demand-ing in many ways (e.g. movement artifacts). First, the SCG and ECG data were processed by filter mechanisms (Butterworth 3rd order [0.25Hz, 85Hz]) and annotated by peak-detection based on patter-analysis. With the anno-tated data, a detailed examination of the established SCG features with regard to absolute and relative amplitudes and time intervals was realized. Through this analysis it is possible to make initial comparisons of the child groups to adult data. The infant's SCG amplitude is up to five times smaller, but the actual signal morphology is equal. Established physiological process-es of the child's heart can also be identified (e.g. high pulse-rate-variability). This work is a prelude to a further collection and should lead to an open-data database that includes data a wide range of ages as well as various patholo-gies.