Patient-Specific Lumped-Parameter Model for Quantifying Vessel-Specific Remodeling and Predicting Right Ventricular...

Area of Science:

• Cardiovascular Physiology
• Computational Modeling
• Medical Imaging

Background:

• Pulmonary hypertension (PH) is a complex disease characterized by patient-specific variability and vessel remodeling, often leading to right ventricular (RV) failure.
• Current gold standard for RV assessment, pressure-volume (PV) loop acquisition, is invasive and restricted to specialized clinical settings.
• There is a need for noninvasive methods to assess RV function and quantify vascular remodeling in PH.

Purpose of the Study:

• To develop a patient-specific lumped-parameter model for quantifying vessel-specific remodeling in PH.
• To simulate RV PV loops across different PH phenotypes using routine clinical data.
• To assess the potential of computational models as noninvasive tools for PH phenotyping and RV function assessment.

Main Methods:

• A lumped-parameter model was developed and calibrated using data from right heart catheterization and echocardiography.
• Model performance was evaluated by R-squared values for pressure and flow, and by comparing derived hemodynamic metrics with clinical values.
• Dimensionality reduction techniques, including linear discriminant analysis (LDA), were employed to assess the separability of PH phenotypes based on model-derived features.

Main Results:

• The lumped-parameter model demonstrated good agreement with clinical data across the PH cohort.
• Model-derived vessel-specific parameters revealed distinct physiological differences among PH phenotypes.
• Simulated RV PV loops showed phenotype-specific variations in RV volumes, pressures, and stroke work, with LDA indicating that model-derived features provide additional discriminatory information.

Conclusions:

• Lumped-parameter models, when calibrated with clinical data, can effectively quantify vessel-specific remodeling and simulate RV PV dynamics.
• These models offer valuable, noninvasive insights for differentiating between various PH phenotypes.
• Computational modeling presents a promising avenue for clinically feasible, in-depth assessment of pulmonary vascular and RV function in PH.