Morphology enabled dipole inversion for quantitative susceptibility mapping using structural consistency between the...

Area of Science:

• Medical Imaging
• Biophysics
• Computational Biology

Background:

• Quantitative Susceptibility Mapping (QSM) is crucial for MRI-based tissue property analysis.
• The conventional QSM problem is ill-posed due to undersampling in the Fourier domain.
• Existing methods often struggle with accuracy and resolution.

Purpose of the Study:

• To develop an improved QSM algorithm for accurate tissue susceptibility mapping.
• To address the ill-posed nature of the QSM problem using structural information.
• To validate the new approach through simulations, phantom experiments, and in vivo imaging.

Main Methods:

• Developed a Morphology Enabled Dipole Inversion (MEDI) approach for QSM.
• Exploited structural consistency between susceptibility and magnitude images.
• Implemented L1 norm minimization to enforce sparsity of edge-based voxels.

Main Results:

• The MEDI approach demonstrated superior performance compared to L2 minimization methods in simulations and phantom studies.
• Numerical simulations and phantom experiments confirmed the effectiveness of L1 minimization.
• Preliminary in vivo brain imaging in healthy subjects and patients with intracerebral hemorrhages showed QSM feasibility.

Conclusions:

• The MEDI algorithm offers a robust solution for the ill-posed QSM problem.
• L1 norm minimization enhances the accuracy of susceptibility mapping by leveraging structural priors.
• QSM is a feasible and promising technique for clinical brain imaging applications.