A Morphology-Enabled Predictor Corrector Algorithm for Gradient Echo MRI Phase Unwrapping

Motivation: Phase wraps in MRI data pose challenges in applications like Quantitative Susceptibility Mapping (QSM) especially at high magnetic field strengths. Goal(s): To develop a more robust and accurate phase unwrapping technique that simultaneously resolves spatial and temporal phase wraps of all echoes in multi-echo gradient echo (GRE) sequences. Approach: The Morphology-Enabled Spatio-Temporal Prediction Correction Unwrapping (MEST-PCU) algorithm is proposed, extending existing methods by incorporating morphological information from magnitude images to guide the unwrapping process. Results: MEST-PCU effectively reduced unwrapping artifacts across multiple datasets of different field strengths and modalities, outperforming conventional unwrapping methods, especially for high-field MRI. Impact: The MEST-PCU method enhances phase unwrapping by integrating morphological information, providing a more accurate and robust solution, particularly at high magnetic field strengths. This method improves QSM estimation even from later echoes, enabling effective susceptibility source separation.