Towards a general framework for fast and feasible k-space trajectories for MRI based on projection methods

Area of Science:

• Medical Imaging
• Biophysics
• Computer Science

Background:

• Reducing scan time in magnetic resonance imaging (MRI) is crucial for clinical applications.
• Non-Cartesian k-space trajectories offer advantages in artifact reduction and motion insensitivity compared to traditional methods.

Purpose of the Study:

• To propose a generalized framework for generating feasible non-Cartesian k-space trajectories.
• To enable the construction of trajectories from both random and structured initial paths, including those based on the traveling salesman problem (TSP).

Main Methods:

• A projection-based framework was developed to generate non-Cartesian k-space trajectories.
• Simulations were performed on phantom and brain MRI images (128x128 and 256x256) using compressed sensing.
• Performance was evaluated using structural similarity (SSIM) index and peak signal-to-noise ratio (PSNR).

Main Results:

• Traveling salesman problem-based trajectories with constant acceleration parameterization (CAP) showed better reconstruction than constant velocity parameterization (CVP).
• Random-like trajectories outperformed TSP-based trajectories in both reconstruction quality and reduced read-out time.
• The proposed projection with permutation (PP) method achieved up to a 67% reduction in read-out time.

Conclusions:

• The proposed generalized framework effectively generates feasible non-Cartesian k-space trajectories.
• Random-like trajectories represent a promising approach for accelerating MRI acquisition while maintaining image quality.
• The projection with permutation (PP) method demonstrates significant potential for reducing MRI scan times.