Non-Cartesian MRI Parallel Imaging Grid Reallocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-Cartesian MRI imaging methods face challenges in high-speed computation and practicality due to complex aliasing artifacts when applying parallel MRI techniques, particularly in radial scan data sampling, which requires a large number of radial scanning lines to prevent reconstruction artifacts.
Innovation Solution
The proposed solution involves reallocating k-space data on orthogonal grid points with a narrower field of view than the target image, using multiple receiver coils with different sensitivity distributions, and performing parallel imaging operations to remove aliasing in real space, thereby simplifying the computation and reducing imaging time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If parallel MRI is applied to non-Cartesian imaging using the generalized parallel method, then aliasing artifacts can be removed, but computation time increases significantly and practicality decreases
Solution Approach 1:
The patent segments the non-Cartesian k-space data into multiple orthogonal sub-k-spaces, each corresponding to a different phase-encoding direction. This segmentation allows the complex non-Cartesian parallel imaging problem to be divided into multiple simpler Cartesian parallel imaging problems that can be solved independently and more efficiently, resolving the contradiction between artifact removal and computation time.
Solution Approach 2:
The patent introduces an intermediary orthogonal k-space representation as a mediator between the non-Cartesian measurement data and the final image reconstruction. By transforming non-Cartesian data into orthogonal k-space components through sensitivity encoding, the method enables the use of efficient Cartesian imaging algorithms while maintaining the benefits of non-Cartesian sampling, thus reducing computation time while removing artifacts.
2Reliability
If the number of radial scanning lines is increased to prevent reconstruction artifacts in non-Cartesian imaging, then image quality improves, but imaging time increases
Solution Approach 1:
The patent segments the radial scanning process into multiple orthogonal Cartesian sampling sets, where each set acquires data along a different orthogonal direction. This segmentation allows the system to achieve complete k-space coverage with fewer total scanning lines by distributing the sampling across multiple orthogonal trajectories, thereby reducing imaging time while maintaining artifact-free reconstruction.
Solution Approach 2:
The patent transitions from single-dimension radial sampling to multi-dimension orthogonal sampling by introducing multiple phase-encoding directions. This dimensional expansion allows the system to cover k-space more efficiently by sampling along multiple orthogonal axes simultaneously, reducing the total number of required scanning lines while preventing reconstruction artifacts.
3Area of stationary object
If multiple receiver coils with different sensitivity distributions are used, then field of view and sensitivity are improved, but device complexity increases
Solution Approach 1:
The patent makes the multiple receiver coil system universal by demonstrating that the same sensitivity distribution information serves multiple functions: it enables both image reconstruction and parallel imaging acceleration, and it works with both Cartesian and non-Cartesian sampling schemes. This multi-functionality justifies the increased device complexity by providing broad applicability and versatility.
Solution Approach 2:
The patent implements self-service by using the sensitivity distribution data acquired during the imaging process itself to automatically perform both the image reconstruction and the parallel imaging operations. The system uses its own coil sensitivity characteristics to simplify the reconstruction process, eliminating the need for separate calibration procedures or additional hardware, thus managing complexity through intelligent self-utilization.
Data Source
AI summary
An MRI apparatus capable of performing a high-speed operation for removing aliasing from the data measured by non-Cartesian imaging in a real space with a small amount of operation is provided. Non-Cartesian data sampling is performed by thinning the number of data by using multiple receiver coils having different sensitivity distribution from each other. Image reconstruction means creates orthogonal data by gridding non-orthogonal data obtained by each receiver coil on a grid having an equal spatial resolution to and a narrower field of view than a target image, subjects it to Fourier transform and creates the first image data containing aliasing components. The second image data is created by using the first image data created for each receiver coil and a sensitivity distribution of each receiver coil.