Dynamic pMRI Reconstruction Using GRAPPA-Operator on Under-Sampled Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Dynamic magnetic resonance imaging (MRI) techniques face challenges in acquiring images quickly while maintaining a sufficient signal-to-noise ratio, as existing methods require long acquisition times for reference data sets, which can be corrupted by motion, especially when using partial parallel acquisition strategies like SENSE and GRAPPA.
Innovation Solution
The GRAPPA-operator technique allows for the computation of reconstruction parameters from fewer under-sampled time frames, using relationships between neighboring k-space lines to fill in missing data and reduce motion-related artifacts, enabling faster and more accurate image reconstruction in dynamic parallel MRI.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional time-interleaved phase encoding is used to achieve acceleration factor R, then image acquisition speed is improved, but reference data set acquisition time increases and motion artifacts increase
Solution Approach 1:
The patent applies partial sampling by acquiring only a subset of k-space lines (undersampling) rather than complete sampling. Multiple under-sampled time frames are acquired instead of a single fully sampled reference frame, reducing the time penalty while still enabling parallel reconstruction through techniques like SENSE or GRAPPA
2Measurement precision
If fully sampled reference data set is acquired to support parallel reconstruction, then reconstruction accuracy is improved, but acquisition time increases and motion corruption increases
Solution Approach 1:
The patent uses partial sampling by acquiring multiple under-sampled time frames instead of a single fully sampled reference frame. The undersampling is compensated through parallel imaging reconstruction techniques that use coil sensitivity information to reconstruct the full image from the partial data
Solution Approach 2:
The patent performs preliminary acquisition of multiple under-sampled time frames that can be assembled into a reference data set. This preliminary data collection is optimized to require less time than conventional fully sampled acquisition, enabling faster setup for parallel reconstruction
3Productivity
If multiple under-sampled time frames are assembled to create reference data set, then parallel reconstruction is enabled, but motion artifacts increase due to long acquisition time
Solution Approach 1:
The patent uses multiple under-sampled time frames with partial k-space coverage instead of a single fully sampled frame. This partial sampling approach reduces the total acquisition time, thereby reducing motion artifacts while still providing sufficient data for parallel reconstruction through techniques like SENSE or GRAPPA
Data Source
AI summary
Example systems, methods, and apparatus facilitate providing a k-space line that is missing in an under-sampled time frame. The missing line is computed by applying a GRAPPA-operator to a known k-space line in the under-sampled time frame. One example method includes controlling a dynamic parallel magnetic resonance imaging (DpMRI) apparatus to acquire a first under-sampled time interleaved frame having at least one first k-space line and controlling the DpMRI apparatus to acquire a second under-sampled time interleaved frame having at least one second k-space line that neighbors the first k-space line. The method includes assembling a reference data set from the first under-sampled time frame and the second under-sampled time frame and then determining the GRAPPA-operator from neighboring k-space lines in the reference data set.