Rapid quantitative MRI at 46 mT: Accelerated T₁ and T₂ mapping with low-rank reconstructions.

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging (MRI)

Background:

• Ultra-low-field MRI offers potential advantages but faces challenges with scan speed.
• Accelerated imaging techniques are crucial for improving the clinical utility of low-field MRI.

Purpose of the Study:

• To evaluate low-rank reconstruction methods for accelerating T1- and T2-mapping in ultra-low-field MRI.
• To assess the performance of image-based locally low-rank (LLR) and k-space-based structured low-rank (SLR) algorithms.

Main Methods:

• Implemented LLR and SLR algorithms on a 46 mT MRI scanner for T1 and T2 mapping.
• Acquired data using 3D turbo spin-echo sequences with variable-density random sampling.
• Validated reconstructions using phantom and in vivo data from healthy volunteers at various undersampling factors (R) and resolutions.

Main Results:

• Both LLR and SLR algorithms successfully reconstructed T1 and T2 maps, reducing scan times significantly (e.g., R=4 reduced T1/T2 mapping time from 45/38 min to 11/9 min).
• Reconstructions eliminated undersampling artifacts, yielding consistent T1 and T2 values across different undersampling factors.
• In vivo experiments showed comparable image quality and T1/T2 values between fully sampled and R=4 undersampled data, consistent with literature.

Conclusions:

• Low-rank reconstruction techniques effectively accelerate T1 and T2 mapping in low-field MRI.
• These methods maintain diagnostic image quality, enhancing the feasibility of quantitative MRI at lower field strengths.