Compressed sensing reconstruction for magnetic resonance parameter mapping.

Area of Science:

• Medical Imaging
• Signal Processing
• Biophysics

Background:

• Magnetic Resonance Imaging (MRI) data acquisition is time-consuming.
• Compressed Sensing (CS) offers acceleration by leveraging signal sparsity.
• Prior knowledge can optimize sparsifying transforms for MRI.

Purpose of the Study:

• To develop and evaluate a Compressed Sensing (CS) reconstruction method for Magnetic Resonance (MR) parameter mapping.
• To utilize a data-driven, overcomplete dictionary for signal sparsification in MR parameter mapping.
• To assess the accuracy of T1 and T2 mapping with significantly reduced MR data.

Main Methods:

• A model-based Compressed Sensing (CS) reconstruction approach was developed.
• An overcomplete dictionary was learned from the data model to sparsify MR signals.
• The method was validated through simulations and in vivo T1 and T2 mapping of the brain.

Main Results:

• Accurate T1 and T2 maps were successfully reconstructed from highly undersampled MR data.
• The model-based CS reconstruction demonstrated effective signal sparsification.
• The approach achieved high-quality parameter maps with reduced acquisition time.

Conclusions:

• Model-based Compressed Sensing (CS) reconstruction enables accelerated MR parameter mapping.
• This technique provides accurate T1 and T2 maps from significantly reduced data.
• The reconstruction framework is adaptable for other MR parameter mapping applications, including diffusion and perfusion imaging.