Sensitivity analysis of an one-dimensional pulse wave propagation model with correlated input

Patient-specific one-dimensional pulse wave propagation models (PWPM) are highly suited for outcome prediction of clinical interventions. Variance- based sensitivity analysis (SA) can be used for parameter-prioritization of these models based on their contribution to the total output uncertainty. Most variance-based SA approaches assume statistical input independence for the sake of computability while in many cases the inputs are correlated which may affect input importance ranking. This study aimed to compare the importance ranking of a SA with and without considering the correlation between model inputs on an one-dimensional PWPM.