Venc Design and Velocity Estimation for Phase Contrast MRI

Area of Science:

• Magnetic Resonance Imaging
• Biomedical Engineering
• Medical Physics

Background:

• Phase-contrast MRI (PC-MRI) measures spin velocity, but phase measurements can be wrapped, complicating velocity estimation.
• Accurate velocity quantification is crucial for diagnosing various cardiovascular and neurological conditions.

Purpose of the Study:

• To introduce a novel, fast, approximate maximum likelihood estimator called Phase Recovery from Multiple Wrapped Measurements (PRoM) for velocity estimation in PC-MRI.
• To enhance the accuracy and range of velocity measurements from multi-coil PC-MRI data, even with amplitude attenuation.

Main Methods:

• PRoM utilizes all pairwise phase differences and their correlations from multi-coil data to minimize velocity estimation error.
• It estimates correlations directly, avoiding the need for coil sensitivity maps or dephasing factors.
• Probabilities of unwrapping errors and velocity estimate distributions are derived for acquisition optimization and post-processing.

Main Results:

• PRoM significantly reduces root mean squared error (e.g., 48.5% decrease in a phantom study) compared to conventional methods.
• The method recovers a wider range of unambiguous velocities, exceeding twice the highest velocity encoding (venc).
• Simulations, phantom, and in vivo studies validate reduced error, increased velocity range, optimized acquisition, and fast computation.

Conclusions:

• PRoM offers a robust solution for accurate velocity estimation in PC-MRI, particularly in challenging scenarios with phase wrapping and dephasing.
• The derived error probabilities enable optimized acquisition strategies and effective post-processing for improved diagnostic confidence.