MRI Motion Correction Using Iterative K-Space Error Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Motion artifacts in magnetic resonance imaging (MRI) due to patient movement during scanning lead to image quality issues, prolonged scanning times, and increased costs, with existing motion correction methods being computationally expensive, requiring additional hardware, or lacking precision.
Innovation Solution
A motion correction method and apparatus that initializes a current motion parameter, calculates a motion-corrected MR image and K-space data, and iteratively updates the parameter based on measurement errors to achieve precise and fast motion correction using fewer parameters, incorporating predefined displacement fields and external motion signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If model-based reconstruction with nonlinear solver is used to correct rigid body motion, then motion correction precision is improved, but computational cost and processing time increase significantly
Solution Approach 1:
The patent segments the motion correction process into two distinct stages: a fast iterative reweighting stage for initial correction and a precise nonlinear solver stage for final refinement. This segmentation allows the computationally intensive nonlinear solver to be applied only when necessary, while the iterative reweighting method handles the bulk of the computation efficiently, thus resolving the contradiction between precision and speed.
Solution Approach 2:
The patent implements a dynamic computational strategy where the choice of reconstruction method (iterative reweighting vs. nonlinear solver) is adapted based on the motion correction requirements and available computational resources. The system dynamically adjusts the correction approach during the reconstruction process, switching between fast approximation and precise calculation as needed, thereby optimizing both speed and precision.
2Adaptability or versatility
If GRICS with external motion signal is used for non-rigid body motion correction, then motion correction capability is improved, but the number of unknown parameters increases making the problem difficult to solve
Solution Approach 1:
The patent extracts and separates the motion correction problem from the image reconstruction problem by using an external motion signal that independently describes the motion state. This external signal serves as a constraint that reduces the number of unknown parameters in the reconstruction process, as the motion parameters are provided by the external signal rather than being estimated from the MR data alone, thus resolving the complexity issue.
Solution Approach 2:
The patent introduces an external motion signal as an intermediary that bridges the gap between the MR imaging data and the motion correction process. This intermediary signal provides motion information that simplifies the reconstruction problem by reducing the number of unknowns, while still enabling comprehensive motion correction for non-rigid body motions.
3Productivity
If FIDNAV or pilot tone signal is used to obtain motion information, then obtaining speed and cost are improved, but detection precision is insufficient for actual motion detection
Solution Approach 1:
The patent implements a feedback mechanism where the motion information obtained from fast methods (FIDNAV or pilot tone signal) is used to initialize the iterative reweighting process, and then refined through iterative correction using the actual MR data. This feedback loop allows the system to leverage the speed of fast methods while achieving the precision of data-driven correction, resolving the contradiction between speed and precision.
Solution Approach 2:
The patent uses fast motion estimation methods (FIDNAV or pilot tone signal) to perform preliminary motion detection and parameter estimation before the main reconstruction process. This preliminary action provides an initial guess that guides the subsequent iterative refinement, allowing the system to benefit from both the speed of fast methods and the precision of detailed correction.
4Reliability
If prospective motion correction with real-time motion parameter detection is used, then motion artifact reduction is improved, but additional hardware or significant sequence change is required prolonging collection time
Solution Approach 1:
The patent performs motion parameter detection during the imaging process itself, using the MR signal data that is already being collected. This preliminary action of extracting motion information from the imaging data eliminates the need for separate detection hardware or additional sequence time, as the motion correction is prepared concurrently with image acquisition, thus resolving the time loss issue.
Solution Approach 2:
The patent enables the MR imaging system to self-detect and self-correct motion artifacts using its own imaging data. The system uses the MR signals already acquired during scanning to estimate motion parameters and perform correction, eliminating the need for external camera systems or navigator modules, and avoiding additional hardware costs and time requirements.
Data Source
AI summary
A motion correction method may include: calculating a current motion-corrected MR image based on a current motion parameter of an imaging target and K-space measurement data of the imaging target; calculating current motion-corrected K-space data based on the current motion parameter of the imaging target and the current motion-corrected MR image; calculating a current K-space measurement data error based on the K-space measurement data of the imaging target and the current motion-corrected K-space data; and determining, based on the current K-space measurement data error, whether an iteration end condition is met. If so, using the current motion-corrected MR image as a final motion-corrected MR image to be used. Otherwise, updating the current motion parameter of the imaging target based on the current K-space measurement data error and the current motion-corrected MR image. The method advantageously provides an increased motion correction speed of an MR image.