Abstract
Seismocardiography signals (SCG) are acoustic vibrations generated by heart activity and measured non-invasively on the surface of the chest. SCG may be used to diagnose and monitor cardiovascular conditions. The signal variability may limit the potentially high SCG clinical utility. It is known that breathing can cause variability, yet it is not well understood. The objective of this study is to quantify SCG and heart rate changes during normal breathing and breath holding (BH). Seismocardiography (SCG), electrocardiography (ECG), and airflow signals were recorded in eight healthy subjects during normal breathing and breath holding (at end inspiration and end expiration). The SCG events were detected and segmented. The heart rate was calculated using the R peak of ECG. Unsupervised machine learning (K-medoid clustering) was implemented using a dynamic time warping (DTW) distance to separate normal breathing SCG waveforms into two clusters. The SCG intra-group variability was calculated in the time domain. Normalized SCG energy in the 0–20 Hz range was also investigated. Results showed that the SCG average intra-cluster variability was 32% lower during breath holding compared to normal breathing. In addition, the average heart rate was 8% lower and normalized SCG energy was 9% lower in breath holding than normal breathing. Variable airflow and lung volume during normal breathing may cause these findings. SCG waveforms during breath holding can be more accurate due to the decreased variability. Hence, it may be useful to collect SCG during breath holding. The results of this study need to be verified with further investigation on a larger number of subjects.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have