MRI EPI N/2 Artifact Suppression via Delayed Signal Averaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current echo planar imaging (EPI) methods in MRI are prone to generating N/2 artifacts due to phase errors in MR signals, which are difficult to completely suppress using existing techniques, especially those involving linear phase corrections and complex addition of signals with opposite polarities, leading to incomplete removal of artifacts and potential image blurring.
Innovation Solution
The proposed MRI apparatus employs a configuration that sets two pulse sequences with different delay times for acquiring MR signals, averaging these signals to generate a combined dataset, and reconstructing images, which can be performed in various data spaces (k-space, hybrid space, or real space) to effectively reduce N/2 artifacts by smoothing pixel values across multiple datasets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional EPI methods are used to acquire MR signals quickly, then imaging speed is improved, but N/2 artifacts occur due to phase errors in the MR signals
Solution Approach 1:
The patent divides the acquisition of MR signals into multiple separate acquisitions with different delay times. Instead of acquiring all signals in a single EPI sequence, the method performs multiple EPI sequences with varying delay parameters, then combines the results through averaging to suppress artifacts while maintaining fast imaging capability
Solution Approach 2:
The patent changes the delay time parameter across multiple EPI acquisitions. By varying this temporal parameter between sequences and combining the data, the method suppresses phase errors that cause N/2 artifacts while preserving the rapid imaging advantage of EPI
2Object-generated harmful factors
If linear phase corrections are applied to suppress N/2 artifacts, then artifact reduction is achieved, but complete removal is difficult and image blurring may occur
Solution Approach 1:
The patent applies phase correction in advance during the data acquisition and preprocessing stage, before final image reconstruction. By correcting phase errors early in the multi-acquisition process and combining corrected datasets through averaging, the method achieves more complete artifact suppression without degrading image quality
Solution Approach 2:
The patent combines multiple MR signal datasets acquired with different delay times through averaging. This merging of multiple corrected datasets suppresses N/2 artifacts more effectively than single-correction methods while maintaining or improving image quality through the combined signal information
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 simplifies the suppression of N/2 artifacts by averaging MR signals acquired with different delay times, improving the signal-to-noise ratio and reducing ghost images, even when nonlinear phase errors are present, without inverting the polarity of readout gradient pulses, thus avoiding image blurring.
Implementation Method 1
a static magnetic field magnet configured to generate a static magnetic field
Implementation Method 2
a gradient coil configured to generate a gradient magnetic field
Implementation Method 3
an RF coil configured to apply a radio frequency pulse to the object and receive a magnetic resonance signal from the object
Data Source
AI summary
An MRI apparatus includes a scanner configured to apply an RF pulse to an object and processing circuity configured to: set a first pulse sequence in which acquisition of a first set of MR signals is started after a first delay time from application of a first excitation pulse, and a second pulse sequence in which acquisition of a second set of MR signals is started after a second delay time from application of a second excitation pulse, the second delay time being different from the first delay time; acquire first and second sets of MR signals by causing the scanner to apply the first and second pulse sequences to the object; generate a combined dataset by averaging a first dataset based on the first set of MR signals and a second dataset based on the second set of MR signals; and reconstruct an MR image based on the combined dataset.