Parallel MRI Unfolding via Extended Coil Sensitivity Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnetic resonance imaging (MRI) systems using parallel imaging face challenges in reducing unfolding artifacts caused by spatially misregistered signals, particularly in echo planar imaging (EPI), which result in distortions and incorrect signal placement in diagnostic images.
Innovation Solution
The technique involves processing coil sensitivity maps by extending them to accurately reflect actual coil sensitivities and generating masks from preliminary unfoldings of main scans to improve image reconstruction, reducing artifacts without increasing acquisition times or introducing noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If parallel imaging is used to reduce acquisition time, then productivity is improved, but unfolding artifacts increase due to spatial misregistration
Solution Approach 1:
The patent applies preliminary action by performing a prescan to generate coil sensitivity maps before the actual imaging scan. These sensitivity maps are then used during the parallel imaging reconstruction process to accurately unfold the aliased images. The prescan also generates a distortion map that is used to correct spatial misregistration, ensuring that the preliminary preparation enables accurate reconstruction without compromising image quality.
Solution Approach 2:
The patent uses an intermediary approach by introducing a distortion map as a mediator between the raw parallel imaging data and the final reconstructed image. This distortion map, generated from the prescan, corrects the spatial misregistration that occurs during parallel imaging. The intermediary distortion correction step allows the system to maintain both fast acquisition and high image accuracy.
2Measurement precision
If sensitivity maps are extended to cover the entire field of view, then measurement precision is improved, but noise is introduced
Solution Approach 1:
The patent applies local quality by differentiating between regions inside and outside the object. The mask generated from the prescan identifies the object boundary, and sensitivity maps are only extended and used within this boundary. This localized approach ensures that sensitivity information is applied only where it is valid and useful, preventing noise from being introduced in regions outside the object where sensitivity extrapolation would be inaccurate.
Solution Approach 2:
The patent uses copying by creating a mask that replicates the object boundary from the prescan data. This mask is then used to constrain the application of extended sensitivity maps during the main scan reconstruction. The copied boundary information from the prescan serves as a template for determining where sensitivity map extension is appropriate, avoiding noise introduction in regions where copying would create artifacts.
3Manufacturing precision
If distortion correction is applied to EPI images, then manufacturing precision is improved, but acquisition time increases
Solution Approach 1:
The patent applies preliminary action by performing distortion correction during the prescan phase rather than during the main imaging scan. The distortion map is generated from the prescan and then applied during the parallel imaging reconstruction. This preliminary preparation of distortion correction parameters enables fast acquisition during the main scan while still achieving accurate spatial correction in the final image.
Solution Approach 2:
The patent merges the distortion correction process with the parallel imaging reconstruction process. Instead of performing distortion correction as a separate step after acquisition, the distortion map generated from the prescan is integrated into the reconstruction algorithm. This merging of distortion correction with the parallel imaging unfolding process eliminates additional acquisition time while maintaining spatial accuracy.
Data Source
AI summary
A magnetic resonance imaging (MRI) system, method and/or computer readable medium is configured to effect improved parallel MR imaging with reduced unfolding artifacts by using either or both of:(a) an unfolded “intermediate” diagnostic image to create a more accurate mask for use in further processing raw image data for final unfolded diagnostic images; and/or(b) an extension of coil sensitivity maps by replication (rather than curve-fitted extrapolation) for use in final unfolding of diagnostic images.