Silent 3D MR sequence for quantitative and multicontrast T1 and proton density imaging.

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Medical Physics
• Biomedical Engineering

Background:

• Traditional MRI sequences can be lengthy and generate significant acoustic noise.
• Accurate T1 and proton density (PD) mapping are crucial for quantitative MRI analysis.
• Bias-field artifacts can affect the accuracy of MRI parameter quantification.

Purpose of the Study:

• To develop a silent, fast, and 3D MRI method for simultaneous T1 and PD mapping.
• To generate bias-corrected, time-series T1-weighted (T1w) images.
• To improve patient comfort and procedure acceptance through reduced acoustic noise.

Main Methods:

• Acquisition of undersampled T1w images at various inversion times (TIs) using a RUFIS sequence with an interleaved k-space trajectory.
• Reconstruction of unaliased images via temporal subspace constraint based on a learned signal model.
• Parameter mapping by fitting data to the signal model and applying bias-field correction to T1w images.

Main Results:

• Phantom studies showed high consistency with gold standard T1 values (R² = 0.9976) and low coefficients of variation (CVs 0.09%-0.83%).
• Volunteer studies yielded T1 values for gray and white matter consistent with literature, with clear histogram delineations (CVs 0.01%-2.30%).
• Measured acoustic noise increase was minimal (2.6 dBA above background).

Conclusions:

• The proposed method enables accurate and repeatable T1 and PD mapping using sparsely sampled T1w images.
• The technique significantly reduces acoustic noise, enhancing patient comfort and MRI procedure acceptance.
• This silent, fast, 3D MRI approach offers a promising advancement in quantitative imaging.