SAMER Scout Scan for MRI Motion Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigator-free retrospective motion correction techniques for magnetic resonance image data are computationally demanding and require large amounts of training data, making them inefficient for clinical use due to the need for repeated updates and instability in estimating hundreds of temporal motion parameters.
Innovation Solution
The implementation of a single rapid scout scan using the SAMER technique, which reduces computation costs and eliminates the need for pre-trained Machine Learning networks, allowing for real-time motion estimation and correction without repetitive updates, thereby stabilizing the motion estimation process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If navigator-free retrospective motion correction techniques are used to estimate hundreds of temporal motion parameters, then motion artifacts can be reduced, but the computational complexity and reconstruction time become excessively high
Solution Approach 1:
The motion correction process is segmented into two independent stages: (1) a single rapid scout scan to obtain an initial motion-free image estimate, and (2) shot-dependent motion parameter estimation using this initial estimate. This segmentation avoids the need to simultaneously estimate hundreds of coupled temporal motion parameters, thereby reducing computational complexity while maintaining motion correction accuracy.
Solution Approach 2:
A preliminary rapid scout scan is performed to obtain an initial motion-free image estimate {circumflex over (x)} before the main motion correction process. This preliminary action provides a stable reference that eliminates the need for repeated updates and iterative refinement, directly reducing the computational burden of the subsequent motion parameter estimation.
2Productivity
If Machine Learning networks are used to obtain initial image estimates for motion correction, then reconstruction speed can be improved, but large amounts of training data are required and repeated updates are needed
Solution Approach 1:
Instead of using expensive, pre-trained Machine Learning networks that require large training datasets and repeated updates, the invention uses a simple, rapid scout scan to obtain an initial motion-free image estimate. This disposable, lightweight approach achieves the same speed-up benefit without the need for extensive training data or iterative refinements.
Solution Approach 2:
The system uses its own rapid scout scan capability to generate the initial image estimate {circumflex over (x)} needed for motion correction, rather than relying on externally trained Machine Learning models. This self-service approach eliminates the dependency on large training datasets and complex pre-training procedures.
3Reliability
If all shots are coupled in the inverse problem for motion estimation, then comprehensive motion correction can be achieved, but the non-convex estimation becomes computationally demanding
Solution Approach 1:
The coupled inverse problem is segmented by using the scout scan-derived initial estimate {circumflex over (x)} to decouple the motion parameter estimation into shot-dependent problems. This allows each shot's motion parameters to be estimated independently and in parallel, dramatically reducing reconstruction time while maintaining estimation accuracy through the stability provided by the initial estimate.
Data Source
AI summary
In a method and system for reducing motion artifacts in magnetic resonance image data, a scout scan of the region of the patient is performed, a magnetic resonance (MR) measurement of the region of the patient is performed to acquire MR image data of the region of the patient, and motion correction is performed on the acquired MR image data based on the scout scan to generate corrected MR image data. The motion correction technique advantageously reduces an influence of a patient motion on the magnetic resonance image data.