Multi-site seismocardiographic measurements for the estimation of cardiac output in patients with aortic stenosis

Cardiac Output (CO) measurement is critical for assessing cardiovascular efficiency, particularly in patients with aortic stenosis undergoing transcatheter aortic valve implantation (TAVI). Traditional CO monitoring techniques, including invasive and non-invasive methods, often provide only intermittent measurements, limiting their ability to track rapid hemodynamic changes. This study presents a novel approach using wearable wireless inertial measurement units (IMUs) to continuously measure CO through electrocardiogram (ECG) and seismocardiogram (SCG) signals. Our method leverages seismocardiographic techniques combined with machine learning algorithms to provide real-time insights into hemodynamic status. Clinical trials involving 11 patients demonstrated the feasibility and accuracy of this approach, with data collected on the patient's chest from multiple auscultation areas, including the aortic, mitral, pulmonary, and tricuspid valves, as well as the xiphoid process. To assess whether the choice of valve position or measurement axis significantly affected the accuracy of CO estimations, a Kruskal–Wallis non-parametric analysis of variance (ANOVA) for each performance metric was performed. This test was chosen due to the small sample size and the lack of normality in the data distributions. The analysis did not reveal any statistically significant difference. This study validates the proposed method's accuracy against the indirect Fick method and investigates different measurement sites. Our findings suggest that continuous, non-invasive CO monitoring using SCG and ECG signals could enhance clinical decision-making during TAVI and potentially other surgical procedures, offering a practical and portable alternative to traditional invasive monitoring systems.