MRI k-space undersampling using orthogonal sampling masks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current MRI scanning technologies face challenges in reducing scan time while maintaining image quality, as they often require continuous sampling along fixed physiological directions, leading to coherence in the sampled data and resulting in blurring and artifacts.
Innovation Solution
The method involves using multiple sampling masks to acquire MR k-space data in orthogonal directions, introducing incoherence into the sampled data, which allows for improved image reconstruction and reduced blurring without increasing scanning time. This is achieved by partitioning 2D or 3D sampling masks to sample data in multiple directions, such as X, Y, and Z, and reconstructing images from the combined data sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If compressed sensing is used to reduce scan time by measuring fewer Fourier coefficients, then acquisition time is reduced, but the readout line remains coherent in one direction which limits the incoherence needed for optimal compressed sensing
Solution Approach 1:
The patent extends the sampling approach from one-dimensional readout lines to two-dimensional sampling patterns in k-space. By sampling along multiple directions (e.g., horizontal, vertical, and diagonal lines) rather than solely along fixed readout lines, the method introduces incoherence in additional dimensions while maintaining the accelerated acquisition enabled by compressed sensing.
Solution Approach 2:
The patent divides the k-space sampling into multiple separate sampling patterns or masks, each sampling along different directions. Instead of using a single coherent readout direction, the overall sampling pattern is segmented into multiple directional components that are combined to achieve the desired incoherence for compressed sensing reconstruction.
2Measurement precision
If non-Cartesian sampling patterns are used to achieve incoherent sampling, then image quality improves by reducing blurring and artifacts, but computation time increases due to non-uniform FFT requirements
Solution Approach 1:
The patent applies different sampling densities and patterns to different regions of k-space. Central regions may be sampled more densely while peripheral regions use sparser sampling patterns. This local variation in sampling quality allows the use of efficient Cartesian or near-Cartesian sampling methods in regions where they are most effective, reducing the need for computationally intensive non-uniform FFT operations while maintaining image quality.
3Speed
If continuous readout lines are acquired sequentially one sample at a time, then acquisition of each line is fast, but skipping samples within a single readout line does not save much time compared to switching to different readout lines
Solution Approach 1:
The patent merges multiple one-dimensional sampling patterns into a unified two-dimensional sampling approach. By combining samples from multiple directions and orientations into a single k-space dataset, the method achieves more effective undersampling than could be obtained by sequential line acquisition alone, improving overall sampling efficiency while maintaining hardware compatibility.
Data Source
AI summary
A method for magnetic resonance (MR) imaging is provided. A first sampling mask is provided for sampling along a first set of parallel lines extending in a first direction in k-space. A second sampling mask is provided for sampling along a second set of parallel lines extending in a second direction in k-space. The second direction is orthogonal to the first direction. A first set of MR k-space data is sampled using an MR scanner, by scanning a subject in the first direction using the first sampling mask. A second set of MR k-space data is sampled using the MR scanner, by scanning the subject in the second direction using the second sampling mask. An MR image is reconstructed from a combined set of MR k-space data including the first set of MR k-space data and the second set of MR k-space data.