MRI EPI Artifact Suppression via Pre-Phase Gradient Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional magnetic resonance echo planar imaging (EPI) technologies face challenges in simultaneously suppressing Nyquist and acceleration artifacts, which affect image quality due to factors like eddy currents and gradient coil heating, and existing methods struggle to effectively address both artifacts simultaneously.
Innovation Solution
The method involves acquiring multiple k-space data sets using imaging sequences with pre-phase-dispersion gradient pulses and phase encoding gradients, where the pre-phase-dispersion gradient pulses have a standard area difference, and averaging the amplitudes of reconstructed magnetic resonance images to generate an average amplitude image, effectively suppressing both Nyquist and acceleration artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional echo planar imaging is used for fast acquisition, then imaging speed is improved, but Nyquist artifacts and acceleration artifacts occur
Solution Approach 1:
The imaging process is segmented into multiple separate acquisitions with different pre-phase-dispersion gradient areas rather than a single acquisition. By dividing the imaging into N separate k-space data sets with progressively increasing gradient areas, the artifact suppression is achieved through segmented sampling in k-space, allowing subsequent averaging to reduce artifacts while maintaining fast imaging capabilities
Solution Approach 2:
The method employs periodic action by acquiring multiple k-space data sets with pre-phase-dispersion gradient pulses that have periodically increasing areas. The gradient areas follow a periodic pattern (0, ΔA, 2ΔA, ..., (N-1)ΔA) across N acquisitions, creating periodic phase variations that enable artifact suppression through averaging while maintaining the fast EPI acquisition rhythm
2Manufacturing precision
If multiple imaging sequences with different pre-phase-dispersion gradient areas are used, then artifact suppression is improved, but acquisition time increases
Solution Approach 1:
The method changes the area parameter of the pre-phase-dispersion gradient pulse across multiple acquisitions rather than changing other sequence parameters. By systematically varying only the gradient area (A, A+ΔA, A+2ΔA, ...) while keeping all other sequence parameters constant, the patent achieves artifact suppression through parameter modulation without requiring substantial increases in acquisition time, as the EPI fast acquisition rhythm is maintained
Solution Approach 2:
The method uses partial action by acquiring only the necessary number of k-space data sets (N sequences) with incremental gradient area changes rather than exhaustive sampling. The standard area difference is specifically set to 2/N of a phase encoding gradient area, which is the minimum required to achieve effective artifact suppression through averaging, avoiding excessive acquisitions that would unnecessarily extend scan time
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach attenuates artifacts in the average MR image, maintaining signal intensity and achieving a good signal-to-noise ratio (SNR) while effectively reducing Nyquist and acceleration artifacts.
Implementation Method 1
a gradient coil, configured to generate an encoding gradient
Implementation Method 2
a radio frequency (RF) coil, configured to generate an RF pulse
Implementation Method 3
each imaging sequence comprising a pre-phase-dispersion gradient pulse and a plurality of phase encoding gradients
Data Source
AI summary
A magnetic resonance imaging system and method, and a computer-readable storage medium are provided. The magnetic resonance imaging method includes: acquiring a plurality of k-space data sets by using a plurality of imaging sequences, each imaging sequence comprising a pre-phase-dispersion gradient pulse and a plurality of phase encoding gradients applied after the pre-phase-dispersion gradient pulse, wherein the pre-phase-dispersion gradient pulses of the plurality of imaging sequences have a standard area difference therebetween when ordered according to area values; respectively reconstructing magnetic resonance images from the respective k-space data sets; and averaging amplitudes of the magnetic resonance images to generate a magnetic resonance image of an average amplitude.