In Silico Investigation of a Mathematical Model Relating the Ballistocardiogram to Aortic Blood Pressure.

Abstract

To investigate the plausibility of a mathematical model of the ballistocardiogram (BCG) called the "BCG-blood pressure (BP) model" developed in our prior work. We conducted a rigorous in silico investigation of the BCG-BP model to assess the hypothesis that the BCG originates from the aortic BP gradients using a mathematical model of systemic arterial circulation that can simulate BP and blood flow wave propagation and reflection in 55 arteries in the arterial tree and the left heart. The BCG-BP model employed the aortic BPs of the systemic arterial circulation model to predict the BCG close in morphology to the ground truth BCG computed as the change in the momentum of the blood in all arteries simulated by the systemic arterial circulation model. The root-mean-squared error and the r value between the two were 4.1% (3.5%-6.8%) and 0.978 (0.958-0.984) (median (IQR)). The errors in the BCG wave timings and normalized amplitudes were adequately small (timing errors<14% and amplitude errors<16% (median)). The relationship between the BCG and aortic BPs implied by the BCG-BP model may be plausible. The BCG is receiving increasing interest for ubiquitous cardiovascular health and disease monitoring by virtue of its ultra-convenient measurement. However, the detailed physiological meaning of the BCG and its relationship to arterial hemodynamics is unknown. The findings from this work can lead to new opportunities for the BCG as a window to view the alterations in aortic BP due to changes in cardiovascular health and disease.