A new phase correction method in NMR imaging based on autocorrelation and histogram analysis

Area of Science:

• Medical Imaging
• Statistical Signal Processing
• Nuclear Magnetic Resonance (NMR)

Background:

• Phase distortions are a common artifact in NMR imaging, degrading image quality and complicating analysis.
• Existing phase correction methods often require specific adjustments or additional data acquisition, limiting their applicability.

Purpose of the Study:

• To introduce a new, versatile statistical approach for phase correction in NMR imaging.
• To demonstrate the effectiveness of the proposed method across various NMR imaging techniques.

Main Methods:

• A two-step statistical correction involving first-order and zero-order phase adjustments.
• Estimation of first-order phase error using image autocorrelation.
• Extraction of zero-order correction from the phase distribution histogram of the corrected image.

Main Results:

• The proposed method successfully corrects phase distortions in NMR images.
• Experimental results show improved image quality for inversion recovery and quadrature spectroscopic imaging.
• The algorithm operates in the spatial domain post-acquisition, requiring no prior adjustments.

Conclusions:

• The developed statistical phase correction algorithm is effective and broadly applicable to phase-sensitive NMR techniques.
• This method offers a practical solution for enhancing NMR image quality without increased acquisition complexity.
• The approach is suitable for techniques such as inversion recovery, quadrature modulated, spectroscopic, and flow imaging.