Low-rank iterative infilling for zero echo-time (ZTE) imaging.

Area of Science:

• Magnetic Resonance Imaging
• Image Reconstruction
• Signal Processing

Background:

• Zero Echo Time (ZTE) imaging is crucial for certain applications.
• Missing data in the dead-time gap of ZTE acquisitions poses reconstruction challenges.
• Existing methods struggle with artifact reduction in ZTE dead-time gaps.

Purpose of the Study:

• Introduce a new referenceless low-rank reconstruction technique.
• Address and overcome limitations caused by missing samples in the ZTE dead-time gap.
• Enhance the quality and reliability of ZTE imaging.

Main Methods:

• Reformulate missing sample in-filling as a low-rank constrained inverse problem.
• Evaluate performance using Monte Carlo simulations and in vivo experimental data.
• Compare the proposed method against algebraic and parallel imaging techniques.

Main Results:

• The technique demonstrates superior performance across various signal-to-noise ratios (5-20 dB) and dead-time gaps up to 4.5 Nyquist dwells.
• Artifact-free reconstruction is achieved for dead-time gaps up to 4 Nyquist dwells.
• Convergence speed decreases exponentially with increasing dead-time gap size.

Conclusions:

• The proposed referenceless low-rank reconstruction method enables artifact-free ZTE imaging.
• It supports higher imaging bandwidths by effectively managing dead-time gaps.
• Outperforms conventional algebraic and parallel imaging methods in ZTE reconstruction.