Synthetic pulse wave dataset for analysis of vascular ageing in elderly patients

This paper presents a methodology to generate synthetic pulse wave database. Each virtual subject is generated with the help of one-dimensional hemodynamics model of systemic circulation with lumped model of the left heart. This paper describes and compares two parameter optimization methods: unscented Kalman filter and Bayesian optimization. As a case study, an experiment is conducted to predict cardio-ankle vascular index (CAVI) values for real individuals with a machine learning algorithm trained on a synthetic population. The average error of 6.5% is achieved