Seed prioritized unwrapping (SPUN) for MR phase imaging

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging (MRI)
• Image Processing

Background:

• Region-growing methods for phase unwrapping can be lossless but suffer from error propagation and long computation times.
• Existing methods require improvement in balancing robustness and efficiency.

Purpose of the Study:

• To develop and demonstrate an accurate, robust, and efficient region-growing phase unwrapping method for MR phase imaging.
• To address limitations in current phase unwrapping techniques.

Main Methods:

• A Seed Prioritized Unwrapping (SPUN) method was developed, using a subset of seed voxels based on continuity quality.
• The method was evaluated using computer simulations, phantom scans, and in vivo brain and pelvis scans at 3 Tesla.
• Error rates, seed percentages, and calculation times were recorded and compared to Laplacian unwrapping.

Main Results:

• SPUN demonstrated robust unwrapping performance with error rates below 0.4% even at low signal-to-noise ratios (SNR) of 1 dB.
• The method achieved visually minimal errors and fast calculation speeds (<4 min for large datasets) for in vivo data with complex morphology.
• Optimal seed priority numbers were identified for 2D and 3D data.

Conclusions:

• SPUN provides a highly robust and fast region-growing-based phase unwrapping solution.
• The method eliminates the need for tissue masking or segmentation and does not impose limitations on imaging parameters.