MRI Body Motion Correction via Hybrid Space Dimensionality Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current nuclear magnetic resonance (NMR) apparatuses face challenges in correcting body motion-induced image degradation during MRI scans, particularly in three-dimensional data processing, which prolongs the time from imaging to image display and reduces accuracy.
Innovation Solution
A nuclear magnetic resonance apparatus that generates a synthesized signal unaffected by body motion by converting three-dimensional frequency space data and weighting factors into hybrid space data, allowing for faster convolution operations and accurate body motion correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If three-dimensional frequency space data is used for body motion correction, then measurement precision is improved, but loss of time increases due to prolonged calculation time
Solution Approach 1:
The patent transforms the three-dimensional frequency space data into hybrid space data by performing Fourier transformation along the slice encoding direction. This dimensionality change converts the 3D convolution operation into a 2D convolution operation, significantly reducing the computational complexity and calculation time while maintaining the accuracy of body motion position detection
2Reliability
If GRAPPA algorithm is used to calculate duplicate signal, then reliability of body motion correction is improved, but loss of time increases due to time-consuming convolution operation
Solution Approach 1:
The patent applies Fourier transformation to convert the 3D frequency space data into hybrid space data, which reduces the convolution operation from three dimensions to two dimensions. This dimensional reduction maintains the reliability of the GRAPPA algorithm for body motion correction while significantly decreasing the computational time required
3Manufacturing precision
If multi-slice data processing is performed, then manufacturing precision is improved, but loss of time increases due to considerable processing time
Solution Approach 1:
The patent performs Fourier transformation along the slice encoding direction to convert multi-slice three-dimensional frequency space data into hybrid space data. This transformation reduces the computational burden of processing multi-slice data by converting 3D convolution to 2D convolution, thereby maintaining image reconstruction quality while reducing the time from imaging to image display
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 significantly reduces calculation time for body motion detection and correction, enabling quicker image display and maintaining high accuracy in body motion position detection and correction, thus improving the efficiency and quality of MRI imaging.
Implementation Method 1
The nuclear magnetic resonance apparatus generates nuclear magnetic resonance in an inspection object
Implementation Method 2
generates nuclear spins of atoms included in the inspection object placed in a static magnetic field
Implementation Method 3
a receive coil that receives the nuclear magnetic resonance signal
Data Source
AI summary
An image corrected for body motion with high accuracy in a short time when performing retrospective body motion correction on an MRI image is provided, and the time from imaging to image display is reduced. A body motion corrector of an MRI apparatus has a weighting factor calculator that calculates a three-dimensional weighting factor based on signals received by multi-channel receive coils. A processing space converter converts three-dimensional frequency space data of a measurement signal and the three-dimensional weighting factor respectively into hybrid space data and a two-dimensional weighting factor. A synthesized signal calculator calculates a synthesized signal by convolution integration of the hybrid space data and the two-dimensional weighting factor; and a body motion position detector detects a body motion occurrence position in the hybrid space from the hybrid space data and the synthesized signal.