Motion-Corrected MRI Reconstruction via Motion-State Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing MRI motion correction techniques are computationally demanding and require significant reconstruction time, making them impractical for clinical use on standard hardware.
Innovation Solution
A method that compresses motion states by identifying and grouping similar motion parameters, reducing the number of states through redundancy analysis, and using a truncated compression matrix to accelerate image reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If fully retrospective motion correction methods are used to reduce motion artifacts, then image quality is improved, but reconstruction time increases significantly
Solution Approach 1:
The patent segments the motion correction process into two distinct phases: (1) a fast scout acquisition phase that provides initial motion estimates, and (2) a main image acquisition phase that uses pre-computed motion correction operators. This segmentation allows the computationally intensive motion estimation to be performed once on low-resolution data, rather than repeatedly on high-resolution data, thereby maintaining image quality while dramatically reducing reconstruction time.
Solution Approach 2:
The patent performs preliminary motion estimation using a rapid scout acquisition before the main image reconstruction. The motion correction operators are pre-computed based on this preliminary motion data, allowing the main reconstruction to proceed with already-prepared correction parameters rather than computing everything from scratch, thus reducing overall reconstruction time while preserving motion correction quality.
2Device complexity
If motion correction algorithms are implemented on standard clinical computing hardware, then device complexity is reduced, but computation time increases
Solution Approach 1:
The patent changes the resolution parameter of the scout acquisition to be lower than the final image resolution. This parameter change allows motion estimation to be performed on less computationally intensive low-resolution data, enabling standard clinical hardware to handle the computation within acceptable timeframes while still providing accurate motion correction for the high-resolution final image.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
The invention relates to a method for reconstructing a motion-corrected magnetic resonance image of a subject, comprising: providing (101) k-space magnetic resonance data comprised of a plurality of shots, wherein each shot corresponds to an individual motion state of the subject; providing (102) motion parameters related to each motion state; determining (103) redundancies across the motion states of the plurality of shots based on the motion parameters; compressing (104) the plurality of motion states based on the determined redundancies across the motion states; reconstructing (105) the magnetic resonance image (8) from the k-space magnetic resonance data based on the compressed plurality of motion states.