Non Invasive Blood Flow Features Estimation in Cerebral Arteries from Uncertain Medical Data.

Area of Science:

• Biomedical Engineering
• Computational Fluid Dynamics
• Medical Imaging

Background:

• Estimating internal carotid artery pressure non-invasively is challenging due to unknown hemodynamic parameters and complex arterial network simulations.
• Accurate cerebral hemodynamic modeling requires patient-specific data and robust parameter estimation techniques.

Purpose of the Study:

• To develop and validate a non-invasive methodology for estimating internal carotid artery pressure.
• To utilize data assimilation and Ensemble Kalman Filters (EnKF) for parameter identification in cerebral network models.

Main Methods:

• Employed patient-specific blood flow rates from Magnetic Resonance Angiography (MRA) and Magnetic Resonance Imaging (MRI).
• Utilized a reduced-order zero-dimensional compartment model for computational efficiency in blood flow dynamics.
• Applied the EnKF as an optimization tool to solve the inverse hemodynamic problem and identify unknown parameters.

Main Results:

• The methodology successfully estimated parameters in a mathematical model of the cerebral network.
• Predicted blood flow rates in the internal carotid arteries showed good agreement with clinical measurements.
• Predicted systolic and diastolic brachial blood pressures also correlated well with clinical data.

Conclusions:

• The proposed non-invasive method using EnKF is a viable approach for estimating internal carotid artery pressure.
• This technique offers a computationally efficient and tractable solution for inverse hemodynamic problems.
• The findings support the potential clinical application of this methodology for patient-specific hemodynamic assessment.