Accelerated multiple-quantum-filtered sodium magnetic resonance imaging using compressed sensing at 7 T

Area of Science:

• Medical Imaging
• Biophysics
• Magnetic Resonance Imaging

Background:

• Multiple-quantum-filtered (MQF) sodium MRI, including enhanced single- and triple-quantum-filtered imaging of 23Na (eSISTINA), offers insights into restricted sodium, a key clinical biomarker.
• However, traditional eSISTINA sequences face challenges with long acquisition times and suboptimal image quality, limiting clinical use.

Purpose of the Study:

• To develop and implement a novel eSISTINA sequence at 7 Tesla (7T) incorporating compressed sensing (CS).
• To accelerate eSISTINA acquisitions using CS without compromising essential image information or biomarker weighting.

Main Methods:

• A 3D spiral-based sampling eSISTINA sequence was implemented at 7T for CS applications.
• Retrospective undersampling of fully sampled phantom and human subject data was performed, comparing CS reconstructions to nonuniform fast Fourier transform (NUFFT).
• Evaluation metrics included structural similarity (SSIM), signal-to-noise ratio (SNR), sodium compartmental weighting, and quantitative accuracy.

Main Results:

• CS-based reconstructions demonstrated reduced noise and enhanced structural detail in both phantom and in vivo images.
• CS generally outperformed NUFFT in SNR and SSIM, preserving important sodium compartmental weightings.
• Accurate in vivo quantitative estimates were achieved with less than 15% error at a twofold undersampling factor.

Conclusions:

• The study successfully implemented an eSISTINA sequence with incoherent sampling at 7T, leveraging CS for acceleration.
• CS enables twofold acceleration of eSISTINA at 7T, yielding lower noise than NUFFT while preserving structural information and quantitative accuracy.
• Reduced acquisition times through CS significantly enhance the potential clinical utility of MQF sodium MRI.