MODEL-BASED MR PARAMETER MAPPING WITH SPARSITY CONSTRAINT

Area of Science:

• Medical Imaging
• Biophysics
• Computational Science

Background:

• Magnetic Resonance (MR) parameter mapping (T1, T2, etc.) is crucial for tissue characterization.
• Long data acquisition times currently limit the clinical utility of MR parameter mapping.

Purpose of the Study:

• To develop and validate a novel model-based method for accelerated MR parameter mapping.
• To enable direct estimation of MR parameters from highly undersampled k-space data.

Main Methods:

• A new model-based parameter mapping technique utilizing an explicit signal model.
• Imposition of a sparsity constraint on parameter values for efficient estimation.
• Development of an algorithm to solve the parameter estimation problem and analysis using estimation-theoretic bounds.

Main Results:

• The proposed method allows direct estimation of MR parameters from undersampled, noisy k-space data.
• Demonstrated acceleration of T2 brain mapping, illustrating the method's practical performance.
• Performance analysis using estimation-theoretic bounds provides a theoretical foundation.

Conclusions:

• The developed model-based approach significantly accelerates MR parameter mapping.
• This method holds promise for enhancing the clinical applicability of quantitative MR imaging.
• Direct estimation from undersampled data offers a pathway to faster and more efficient tissue characterization.