Wearable sensor driven Cardiac model to derive hemodynamic insights during exercise

Abstract

In this paper, we propose a cardiovascular digital twin platform to simulate the effect of exercise on various cardiac parameters of medical importance. The model incorporates the real-time ECG signal from the body-worn sensors to estimate exercise level and compute cardiac variables like left ventricular dynamics, cardiac output, ejection fraction, mean arterial pressure, etc., of an individual while performing exercises. The novel contribution of this work is to determine the cardiac compliances from the morphology of the single-lead ECG signal and Systemic resistance estimation from the information of the exercise level for a specific subject performing physical activities. Simulation results are produced employing an open-source database (Troika), and it shows medically accepted trends in all the derived cardiac variables. The proposed framework can aid to progressively track the cardiovascular efficiency during exercise for patients with cardiac co-morbidity and act as a personalized therapy guide in a care continuum scenario.