Iterative reconstruction of radially-sampled ³¹P bSSFP data using prior information from ¹H MRI

Area of Science:

• Magnetic Resonance Imaging
• Biophysics
• Medical Imaging

Background:

• Direct phosphorus-31 (³¹P) Magnetic Resonance Imaging (MRI) is crucial for metabolic studies.
• Improving the resolution and accuracy of ³¹P MRI is essential for clinical applications.
• Current methods face challenges with signal-to-noise ratio (SNR) and artifacts.

Purpose of the Study:

• To enhance direct ³¹P MR imaging techniques.
• To develop and evaluate 3D density-adapted radially-sampled balanced steady-state free precession (bSSFP) sequences.
• To assess an iterative reconstruction method using hydrogen (¹H) MRI data for improved spatial distribution of phosphocreatine (PCr).

Main Methods:

• Development of 3D density-adapted radially-sampled bSSFP sequences at 7 Tesla (7T).
• Investigation of three bSSFP gradient schemes to optimize SNR.
• Comparison of conventional gridding reconstruction with an iterative method incorporating an ¹H MRI constraint (binary mask).

Main Results:

• A bSSFP scheme with two point-reflected density-adapted gradients yielded the highest SNR.
• The iterative reconstruction effectively reduced partial volume effects and Gibbs ringing artifacts.
• The iterative method provided a more accurate representation of PCr distribution compared to conventional methods.

Conclusions:

• The developed iterative reconstruction of ³¹P bSSFP data using an ¹H MRI constraint improves imaging accuracy.
• This method is suitable for analyzing regions with sharp tissue boundaries, such as the human calf muscle.
• The enhanced imaging technique offers better representation of real PCr distributions.