Rapid two-step dipole inversion for susceptibility mapping with sparsity priors.

Area of Science:

• Medical Imaging
• Biophysics
• Computational Science

Background:

• Quantitative susceptibility mapping (QSM) reconstructs magnetic susceptibility distributions from MRI data.
• Current QSM algorithms face a speed-accuracy trade-off in solving the ill-posed dipole inversion problem.
• Existing methods often rely on prior information, limiting generalizability.

Purpose of the Study:

• To develop a novel two-step dipole inversion algorithm for QSM.
• To improve the speed and accuracy of QSM reconstruction without using prior information.
• To address the limitations of current state-of-the-art QSM algorithms.

Main Methods:

• A two-step dipole inversion approach was implemented.
• The well-conditioned k-space region was reconstructed using a Krylov subspace solver.
• The ill-conditioned k-space region was solved via constrained l1-minimization, utilizing sparsity constraints.

Main Results:

• The proposed algorithm demonstrated superior performance compared to MEDI and HEIDI.
• Achieved lower root-mean-square error and higher coefficient of determination against COSMOS.
• The method was approximately 50 times faster than existing algorithms.

Conclusions:

• The two-step dipole inversion algorithm offers improved QSM reconstruction quality.
• This approach significantly reduces computation time without compromising accuracy.
• The method provides a viable alternative for rapid and accurate QSM analysis.