3D Cone Trajectory Algorithm for MRI Gradient Waveform Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current three-dimensional magnetic resonance imaging (MRI) techniques face challenges in efficiently covering k-space with radial trajectories, leading to long scan times and low signal-to-noise ratio (SNR) efficiency due to unnecessary twists and anisotropic sampling distributions, which result in artifacts and suboptimal image quality.
Innovation Solution
The development of a three-dimensional cone k-space trajectory design algorithm that optimizes gradient waveforms to minimize twist and ensure uniform sampling density, allowing for non-isotropic fields of view and resolutions, and reusing waveforms to cover multiple cone surfaces efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If radial trajectories are used to cover k-space in three-dimensional MRI, then the imaging coverage is improved, but scan times become long and SNR efficiency decreases due to unnecessary twists and anisotropic sampling distributions
Solution Approach 1:
The patent segments the continuous radial trajectory into discrete conical surfaces in three-dimensional k-space. By dividing the imaging volume into multiple cones with optimized trajectories, the method eliminates unnecessary twists while maintaining complete k-space coverage, thereby reducing scan time without sacrificing imaging coverage.
Solution Approach 2:
The patent transitions from two-dimensional radial trajectories to three-dimensional conical trajectories by adding the temporal dimension to the spatial coverage. This dimensional extension allows uniform sampling density across the entire k-space volume while eliminating redundant twists that plague traditional radial methods, achieving both complete coverage and reduced scan time.
2Volume of moving object
If radial trajectories are used to cover k-space in three-dimensional MRI, then the imaging coverage is improved, but SNR efficiency decreases due to anisotropic sampling distributions
Solution Approach 1:
The patent applies local quality optimization by ensuring uniform sampling density at each conical surface while maintaining different trajectory characteristics for different cones. This localized uniformity across multiple cones achieves global uniformity in k-space coverage, improving SNR efficiency without compromising comprehensive imaging coverage.
Solution Approach 2:
The patent achieves homogeneity in sampling distribution across the entire three-dimensional k-space by designing conical trajectories that uniformly sample each cone surface. This homogeneous sampling eliminates the anisotropic distribution problems of radial trajectories, resulting in improved SNR efficiency and artifact reduction while maintaining complete volumetric coverage.
3Ease of manufacture
If traditional trajectory methods are used, then implementation is straightforward, but artifacts increase and image quality decreases
Solution Approach 1:
The patent replaces the traditional mechanical radial trajectory system with a mathematically optimized conical trajectory system. This substitution eliminates the inherent twists and sampling inefficiencies of radial methods that generate artifacts, while the systematic approach to cone generation maintains ease of implementation through algorithmic precision rather than mechanical complexity.
4Device complexity
If conventional MRI sequences are used, then system simplicity is maintained, but scan times are long and motion correction is suboptimal
Solution Approach 1:
The patent introduces dynamic optimization into the MRI sequence by using time-varying conical trajectories that adapt to the imaging requirements. The systematic variation of cone parameters over time enables faster k-space coverage and improved motion correction capabilities, while the underlying systematic structure maintains relative system simplicity through algorithmic control rather than hardware complexity.
Data Source
AI summary
A method of performing magnetic resonance imaging is provided. Sampling requirements are used to define a three dimensional cone trajectory differential equation. The equation is solved to obtain a starting point. A search is performed by performing a plurality of cycles, where each cycle comprises selecting a point on the cone trajectory, working backward from the starting point to reduce twist, providing a failure value if it is determined that when the twist reaches zero it is not possible to return to the origin at a final velocity of zero, and providing a success value if it is determined that when the twist reaches zero it is possible to return to the origin at a final velocity of zero. A plurality of cycles is performed, where each cycle comprises applying a magnetic resonance image excitation and scanning along the calculated cone trajectory and acquiring a readout.