The use of Fourier-domain analyses for unwrapping phase images of low SNR.

Abstract

We report a new postprocessing procedure that uses Fourier-domain data analyses to improve the accuracy and reliability of phase unwrapping for MRI data of low SNR. The developed method first identifies the Fourier-domain energy peak locations corresponding to different image-domain areas from which a robust measurement of image-domain phase gradients can be obtained even for MRI data of low SNR. The phase-gradient information measured from critical brain regions using the above-mentioned Fourier-domain analysis is then combined with the conventional temporal-domain or spatial-domain phase-unwrapping procedure to remove phase wraps. The developed method was tested with MRI data obtained from 30 healthy adult volunteers, and its performance was quantitatively evaluated. The developed Fourier-domain analysis could robustly quantify image-domain phase gradients even for MRI data with low SNR (e.g., SNR 2). Experimental results show that the Fourier-domain analyses could further reduce phase wrap artifact in data produced by the conventional temporal-domain or spatial-domain phase-unwrapping procedures. Our results demonstrate that the developed phase-unwrapping method can reduce residual phase wraps resulting from conventional procedures in critical brain regions (e.g., near the air-tissue interfaces) and should prove valuable for studies that require accurate measurements of MRI phase values, such as QSM, B0 field mapping, and temperature mapping.