Patient‐specific assessment of cardiovascular function by combination of clinical data and computational model with applications to patients undergoing Fontan operation

The assessment of cardiovascular function is becoming increasingly important for the care of patients with single-ventricle defects. However, most measurement methods available in the clinical setting cannot provide a separate measure of cardiac function and loading conditions. In the present study, a numerical method has been proposed to compensate for the limitations of clinical measurements. The main idea was to estimate the parameters of a cardiovascular model by fitting model simulations to patient-specific clinical data via parameter optimization. Several strategies have been taken to establish a well-posed parameter optimization problem, including clinical data-matched model development, parameter selection based on an extensive sensitivity analysis, and proper choice of parameter optimization algorithm. The numerical experiments confirmed the ability of the proposed parameter optimization method to uniquely determine the model parameters given an arbitrary set of clinical data. The method was further tested in four patients undergoing the Fontan operation. Obtained results revealed a prevalence of ventricular abnormalities in the patient cohort and at the same time demonstrated the presence of marked inter-patient differences and preoperative to postoperative changes in cardiovascular function. Because the method allows a quick assessment and makes use of clinical data available in clinical practice, its clinical application is promising.