Sensitivity Analysis of a Mathematical Model Simulating the Post-Hepatectomy Hemodynamics Response

Area of Science:

• Cardiovascular Physiology
• Medical Modeling
• Surgical Risk Assessment

Background:

• A lumped-parameter model simulates cardiovascular hemodynamics after partial hepatectomy.
• Patient-specific parameter tuning is crucial for predicting portal hypertension (PHT) risk.

Purpose of the Study:

• To conduct a global sensitivity analysis (SA) on a cardiovascular model for partial hepatectomy.
• To identify critical model parameters driving clinical outputs and inform patient-specific calibration.
• To reduce computational costs in SA using polynomial chaos expansion.

Main Methods:

• Global sensitivity analysis (SA) was performed on a lumped-parameter cardiovascular model.
• Polynomial chaos expansion was utilized to optimize computational efficiency during SA.
• Model outputs were constrained to physiologically relevant ranges.

Main Results:

• Sensitivity analysis identified key parameters influencing cardiovascular response and PHT risk.
• Specific parameters were highlighted as critical for patient-specific model calibration.
• The study demonstrated a method to reduce computational burden in SA.
• Physiologically constrained parameter distributions were obtained, enabling virtual population creation.

Conclusions:

• The SA provides insights for improving cardiovascular model parameterization in PHT risk assessment.
• The developed pipeline is applicable to other hemodynamics models for patient-specific insights.
• The creation of a virtual population supports future research in surgical outcomes.