Simple method for MR gradient system characterization and k-space trajectory estimation.

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging (MRI)
• Signal Processing

Background:

• Fast imaging techniques in MRI are crucial for reducing scan times but are susceptible to artifacts.
• Gradient system imperfections lead to deviations between intended and actual k-space trajectories, causing image artifacts.
• Existing artifact reduction methods are often trajectory- and orientation-specific.

Purpose of the Study:

• To develop a universal method for characterizing gradient system behavior.
• To predict actual k-space trajectories for arbitrary scan orientations.
• To reduce image artifacts in fast MRI acquisitions.

Main Methods:

• A single linear time-invariant (LTI) characterization of the gradient system was employed.
• The method utilizes convolution to predict k-space trajectories based on gradient system properties.
• Efficient characterization is achieved by comparing Fourier transforms of input and measured waveforms of a test gradient.
• The technique was validated across spiral, interleaved echo-planar, and 3D cones imaging trajectories.

Main Results:

• The LTI characterization effectively predicted k-space trajectories for tested imaging sequences.
• The proposed method demonstrated significant reduction in reconstructed image artifacts.
• The approach proved effective for various k-space trajectories and scan orientations.

Conclusions:

• A novel, efficient method for characterizing gradient system performance in MRI was presented.
• This technique enables accurate prediction of k-space trajectories, mitigating artifacts in fast imaging.
• The findings offer a generalizable solution for improving image quality in diverse MRI applications.