Non-Cartesian MRI Reconstruction for Real-Time Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current MRI techniques face challenges in achieving real-time imaging with high temporal resolution due to long acquisition times and sensitivity to off-resonance artifacts, particularly with Cartesian encoding schemes, which limit the monitoring of dynamic processes and result in geometric distortions and signal losses.
Innovation Solution
A method combining gradient-echo sequences with non-Cartesian k-space trajectories and regularized nonlinear inverse reconstruction, which estimates coil sensitivities and image content simultaneously, allowing for rapid and motion-robust imaging with enhanced undersampling factors, and implementation on a GPU for reduced reconstruction times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If Cartesian encoding schemes are used for MRI acquisition, then reconstruction is simple and tolerance to instrumental imperfections is high, but acquisition time is long and real-time monitoring of dynamic processes is hampered
Solution Approach 1:
The patent inverts the conventional approach by using non-Cartesian (radial) encoding schemes instead of Cartesian grids. This inversion allows each radial spoke to cross the center of k-space, capturing both high and low spatial frequencies simultaneously, thereby enabling real-time monitoring of dynamic processes while maintaining reconstruction capability through iterative algorithms.
Solution Approach 2:
The patent changes the sampling parameters from Cartesian grid patterns to radial trajectories in k-space. This parameter change transforms the acquisition strategy to capture dynamic information more efficiently, with each radial spoke providing complementary information about moving structures, thus reducing acquisition time for real-time imaging.
2Loss of time
If single-shot gradient-echo sequences are used for high-speed acquisition, then acquisition time is reduced, but geometric distortions and signal losses occur due to sensitivity to off-resonance effects
Solution Approach 1:
The patent applies preliminary action by performing iterative reconstruction that estimates and compensates for coil sensitivities and off-resonance effects before final image formation. The nonlinear inverse reconstruction process pre-corrects for distortions by modeling the physical processes that cause them, thereby eliminating geometric distortions and signal losses in the final image.
Solution Approach 2:
The patent implements feedback through its iterative reconstruction algorithm that continuously refines the image estimate by comparing predicted k-space data with actual measurements. The nonlinear inverse model uses feedback from the data to adjust coil sensitivity estimates and correct for off-resonance effects, progressively improving image quality and reducing artifacts.
3Productivity
If radial data sampling with view sharing is used, then frame rate of 20 images per second is achieved, but image quality deteriorates due to gridding interpolation and sliding window technique
Solution Approach 1:
The patent extracts the problematic gridding interpolation step from the reconstruction pipeline and replaces it with a direct nonlinear inverse reconstruction approach. By taking out the gridding step, the patent eliminates the associated image quality deterioration and artifacts while maintaining the high frame rate capability through efficient iterative algorithms.
Solution Approach 2:
The patent substitutes the mechanical gridding interpolation process with a computational nonlinear inverse reconstruction system. Instead of mechanically interpolating data onto a Cartesian grid, the patent uses a mathematical model that directly reconstructs images from radial k-space data, thereby improving image quality while maintaining productivity.
4Productivity
If high undersampling factors are used to reduce acquisition time, then frame rate increases, but image quality and temporal fidelity deteriorate
Solution Approach 1:
The patent applies dynamics by using a sliding window approach that continuously updates the reconstruction using the most recent data. This dynamic reconstruction strategy maintains temporal fidelity by always incorporating the latest information while allowing high undersampling factors, thereby achieving high frame rates without sacrificing image quality or temporal accuracy.
Solution Approach 2:
The patent ensures continuity of useful action by continuously acquiring and reconstructing images in real-time using overlapping data windows. This continuous process maintains temporal fidelity by ensuring that each frame is reconstructed from the most current available data, while the iterative nature of the reconstruction allows for high undersampling factors to be used.
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 enables continuous acquisition of high-quality MR images with significantly reduced acquisition times, achieving frame rates of up to 50 frames per second and improved image quality by reducing artifacts and geometric distortions, while maintaining robustness to motion and off-resonance effects.
Implementation Method 1
magnetic resonance (MR) images
Implementation Method 2
spatially encodes an MRI signal using a non-Cartesian k-space trajectory
Implementation Method 3
simultaneous estimation of a sensitivity of the at least one receiver coil and the image content
Implementation Method 4
implementation on a GPU for reduced reconstruction times
Data Source
Figure 1~2
Figure 3A~3C
Figure 4A~4D
AI summary
A method for reconstructing a sequence of magnetic resonance (MR) images of an object under investigation, comprises the steps of (a) providing a series of sets of image raw data including an image content of the MR images to be reconstructed, said image raw data being collected with the use of at least one radiofrequency receiver coil of a magnetic resonance imaging (MRI) device, wherein each set of image raw data includes a plurality of data sampies being generated with a gradient-echo sequence, in particular a FLASH sequence, that spatially encodes an MRl signal received with the at least one radiofrequency receiver coil using a non-Cartesian k-space trajectory, each set of image raw data comprises a set of homogeneously distributed lines in k-space with equivalent spatial frequency content, the lines of each set of image raw data cross the center of k-space and cover a continuous range of spatial frequencies, and the positions of the lines of each set of image raw data differ in successive sets of image raw data, and (b) subjecting the sets of image raw data to a regularized nonlinear inverse reconstruction process to provide the sequence of MR images, wherein each of the MR images is created by a simultaneous estimation of a sensitivity of the at least one receiver coil and the image content and in dependency on a difference between a current estimation of the sensitivity of the at least one receiver coil and the image content and a preceding estimation of the sensitivity of the at least one receiver coil and the image content.