Sampling errors in projection reconstruction MRI

Area of Science:

• Medical Imaging
• Image Reconstruction
• Signal Processing

Background:

• Projection reconstruction (PR) techniques in magnetic resonance imaging (MRI) can lead to artifacts.
• These artifacts are often attributed to insufficient k-space sampling, potentially requiring finer sampling than the Nyquist criterion suggests.
• Existing methods may not adequately address the weighting of low spatial frequencies in k-space.

Purpose of the Study:

• To investigate the adequacy of the Nyquist sampling criterion for PR in MRI.
• To develop improved methods for k-space interpolation and filtering to mitigate artifacts.
• To validate the proposed methods through simulations.

Main Methods:

• Analysis of PR techniques, drawing parallels with X-ray computed tomography implementations.
• Development and application of a band-limited interpolation method in k-space.
• Derivation and testing of a suitable k-space filter function to properly weight low spatial frequencies, moving beyond simple 'k' filters.

Main Results:

• The Nyquist sampling criterion is confirmed as adequate for avoiding aliasing in PR, provided band-limited interpolation is employed.
• A novel procedure for constructing an appropriate k-space filter function is described and validated.
• Simulated reconstructions of circular disks demonstrate that the combined methods achieve near-perfect results within the Nyquist limit.

Conclusions:

• Artifacts in MRI PR are avoidable with appropriate k-space interpolation and filtering.
• The Nyquist criterion remains valid for PR when combined with advanced signal processing techniques.
• The developed methods offer a robust solution for high-fidelity image reconstruction in MRI PR.