Personalized Hemodynamic Modeling of the Human Cardiovascular System: A Reduced-Order Computing Model

Area of Science:

• Cardiovascular physiology
• Biomedical engineering
• Computational modeling

Background:

• Personalized hemodynamic modeling is vital for cardiovascular system (CVS) functional prediction.
• Existing reduced-order 1D/0D models estimate hemodynamics but lack practical personalization.
• There is a need for accurate, individualized models to assess CVS function.

Purpose of the Study:

• To present a novel 0-1D coupled, personalized hemodynamic model for the CVS.
• To predict both pressure waveforms and flow velocities in arteries.
• To facilitate patient-specific assessment of physiological and pathological CVS functions.

Main Methods:

• A multiscale CVS model was combined with the Levenberg-Marquardt optimization algorithm.
• An inverse problem was solved using measured blood pressure waveforms.
• Noninvasive measurements of hemodynamic characteristics were collected from 62 volunteers.

Main Results:

• The model achieved a mean square error of 7.1 mmHg² for individual pressure waves.
• Simulated blood flow velocities in the carotid artery correlated well with ultrasound measurements (average R=0.911).
• Estimated arterial stiffness distribution was physiologically realistic.

Conclusions:

• The developed model is efficient and versatile for personalized hemodynamic analysis.
• It accurately fits individual pressure waveforms and reasonably predicts flow waveforms.
• This approach supports individualized patient-specific assessment of CVS functions.