Magnetic Resonance Slice Separation Algorithm k-Space Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Simultaneous multi-slice (SMS) magnetic resonance imaging techniques face a reduced signal-to-noise ratio (SNR) due to the limitations of existing calibration methods, which are not applicable for slice separation algorithms in SMS imaging, leading to SNR penalties.
Innovation Solution
The method involves recording magnetic resonance data in two sub-areas with different sampling degrees, applying a slice separation algorithm to each portion separately, and recombining them to improve the SNR, using a turbo spin echo sequence and slice GRAPPA algorithm, with additional k-space lines sampled to enhance the SNR.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If simultaneous multi-slice imaging with uniform undersampling is used, then imaging speed is improved, but signal-to-noise ratio deteriorates
Solution Approach 1:
The patent segments k-space into a central sub-area and outer areas, applying different sampling strategies to each segment. The central sub-area is completely sampled while outer areas are undersampled, allowing the slice separation algorithm to be calibrated using fully sampled reference data from the central region, thereby improving SNR while maintaining accelerated imaging speed.
Solution Approach 2:
The patent applies local quality by using complete sampling in the central k-space region for calibration purposes while using undersampling in outer regions for acceleration. This local differentiation allows optimal SNR in critical calibration areas while maintaining overall imaging efficiency.
2Measurement precision
If complete sampling of k-space is performed, then signal-to-noise ratio is improved, but imaging time increases
Solution Approach 1:
The patent applies partial sampling by completely sampling only the central sub-area of k-space rather than the entire k-space. This partial complete sampling provides sufficient data for calibrating the slice separation algorithm and determining undersampling patterns, achieving adequate SNR without the time cost of complete full k-space sampling.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach results in an improved signal-to-noise ratio by allowing complete sampling around the k-space center and undersampling in the outer areas, enabling effective slice separation and calibration of the undersampling algorithm, thus enhancing image quality.
Implementation Method 1
Magnetic resonance imaging is an established modality for recording medical image data in examination areas of a patient
Implementation Method 2
a gradient coil arrangement (33) for generating magnetic field gradients in the examination area
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Method for acquiring a magnetic resonance imaging dataset of an examination area of a patient using a magnetic resonance device, wherein a multi-slice imaging technique is used to acquire magnetic resonance data (6, 14, 16) with simultaneous at least partial undersampling in the slice plane, wherein magnetic resonance data (6, 14, 16) are read out simultaneously from several excited slices (5) and assigned to the simultaneously read out slices (5) by a slice separation algorithm, which is calibrated using reference data (1, 1', 3) acquired in a separate reference scan, after which a undersampling algorithm compensating for the undersampling in the slice plane is applied to the undersampled magnetic resonance data (16) of the individual slices (5), wherein the magnetic resonance data (6) are in at least two sub-areas (7,8) of the sampled k-space in the layer plane are recorded with a different sampling rate and, prior to the application of the layer separation algorithm, are split into at least two components (9, 9', 10) each assigned to a sub-area (7, 8) and each of a fixed sampling rate, to which the layer separation algorithm is applied separately, wherein the components (9, 9', 10) are recombined layer by layer to determine the magnetic resonance image data set (15).