K-Space Segmentation for Faster, Higher-Quality MRI Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing k-space undersampling patterns in magnetic resonance imaging (MRI) reduce image quality, interfering with accurate diagnosis and slowing down the imaging process.
Innovation Solution
A system and method that divides the k-space into regions with different sampling patterns, including full, irregular, and regular undersampling, and employs iterative procedures like soft-thresholding and total variation algorithms to enhance image reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If k-space undersampling pattern is implemented to accelerate MRI, then imaging speed is improved, but image quality deteriorates
Solution Approach 1:
The k-space is divided into multiple regions (e.g., central region, intermediate regions, peripheral regions), and different sampling patterns are assigned to different regions. The central region uses full sampling to preserve image quality, while peripheral regions use undersampling to accelerate imaging, thus resolving the contradiction between speed and quality.
Solution Approach 2:
Different sampling densities are applied to different spatial frequencies in the k-space. High-frequency components (peripheral k-space) are undersampled less critically than low-frequency components (central k-space), allowing acceleration while maintaining essential image quality. This local differentiation resolves the uniform quality loss problem.
2Loss of time
If irregular undersampling pattern is applied to accelerate imaging, then scanning time is reduced, but reconstruction accuracy deteriorates
Solution Approach 1:
The system uses feedback mechanisms where the undersampling pattern is designed based on the image reconstruction results, and the sampling strategy is adjusted iteratively to improve reconstruction accuracy while maintaining acceleration benefits.
Solution Approach 2:
Calibration data is acquired beforehand to establish the relationship between undersampled k-space data and final image quality. This preliminary action allows the system to optimize the undersampling pattern in advance, improving reconstruction accuracy without requiring additional scanning time.
Data Source
AI summary
Systems and methods for magnetic resonance imaging acceleration. The systems may perform the methods to obtain imaging data of a subject, or a portion thereof, captured by the MRI system according to an undersampling pattern; execute a first iterative procedure; determine that the first iteration number meets a first threshold; execute, in response to the determination that the first iteration number meets the first threshold, a second iterative procedure; determine that a sum of the first iteration number and the second iteration number meets a second threshold; and generate a reconstruction image of the subject, or a portion thereof, according to the processed imaging data, wherein the first threshold is lower than the second threshold.