Computational assessment of model-based wave separation using a database of virtual subjects

Area of Science:

• Cardiovascular Physiology
• Biomedical Engineering
• Medical Imaging Analysis

Background:

• Arterial wave reflection quantification is crucial in arterial pulse wave analysis.
• Wave separation analysis (WSA) traditionally requires both aortic pressure and flow waveforms.
• Mathematical models are being developed to estimate aortic flow from pressure waveforms alone.

Purpose of the Study:

• To investigate and verify the model-based wave separation (WSA) using the ARCSolver method.
• To assess the accuracy of ARCSolver in estimating arterial wave reflection non-invasively.
• To validate the ARCSolver method on a large virtual cohort of healthy subjects.

Main Methods:

• Utilized an open-access virtual database of 3325 simulated healthy subjects with varied physiological parameters.
• Applied the ARCSolver method to perform WSA using virtual aortic root pressure waveforms.
• Compared ARCSolver results against reference WSA values derived from both pressure and flow waveforms.

Main Results:

• Demonstrated good overall agreement between the model-based ARCSolver method and the reference WSA.
• Reported mean differences for characteristic impedance (-0.05±0.02AU), forward pressure amplitude (-3.93±1.79mmHg), backward pressure amplitude (1.37±1.56mmHg), and reflection magnitude (12.42±4.88%).
• The ARCSolver method showed feasibility as a surrogate for measured flow waveforms.

Conclusions:

• The ARCSolver method provides a reasonable and feasible approach for assessing arterial wave reflection non-invasively in healthy subjects.
• Model-based wave separation using ARCSolver is a viable alternative when flow waveform measurements are unavailable.
• This study supports the potential of ARCSolver for clinical applications in arterial pulse wave analysis.