A new algebraic method for quantitative proton density mapping using multi-channel coil data.

Area of Science:

• Quantitative Magnetic Resonance Imaging (qMRI)
• Neuroimaging
• Biophysics

Background:

• Accurate proton density (PD) determination is vital for quantitative MRI and assessing brain tissue organization.
• Existing methods often rely on the inverse linear relationship between longitudinal relaxation rate (T1) and PD.
• A general framework is needed to model complex relationships and receiver bias fields.

Purpose of the Study:

• To develop and validate a novel method for accurate in vivo proton density estimation in MRI.
• To model receiver bias fields using 3D basis functions for improved accuracy.
• To assess the method's performance under varying noise conditions and its application in patient data.

Main Methods:

• Utilized a generalized framework with 3D basis functions to model receiver bias fields and coil sensitivities.
• Applied the method by parcellating the brain into 30mm cubes and determining optimal basis functions.
• Validated the approach using numerical phantoms with simulated noise and multi-channel (32-element) coil data.

Main Results:

• Achieved a bias close to zero and low noise sensitivity for estimated proton densities.
• Demonstrated a root-mean-square error rate of less than 0.2% in realistic 3D simulations.
• Successfully applied the method to estimate proton density in specific brain structures of MS patients.

Conclusions:

• The novel method provides accurate and robust proton density estimation in MRI, even with noise.
• The use of 3D basis functions effectively models receiver bias fields, enhancing quantitative measurements.
• This technique holds promise for improved neuroimaging analysis, particularly in conditions like multiple sclerosis (MS).