Optimization of quasi-diffusion magnetic resonance imaging for quantitative accuracy and time-efficient acquisition

Area of Science:

• Medical Imaging
• Quantitative MRI
• Diffusion MRI

Background:

• Quasi-diffusion MRI (QDI) is a novel technique modeling diffusion dynamics using the continuous time random walk.
• QDI quantifies diffusion by estimating the diffusion coefficient (D) and a fractional exponent (α), indicating non-Gaussian diffusion signal decay.

Purpose of the Study:

• To optimize b-value selection for rapid clinical acquisition of QDI tensor imaging (QDTI) data.
• To evaluate the impact of maximum b-value, number of shells, and Rician noise on QDTI measures.

Main Methods:

• Optimized clinically appropriate QDTI acquisitions in healthy volunteers.
• Compared acquisitions against a 29-shell multi-b-value reference (MbR) dataset (b=0-5000 s/mm²).
• Investigated effects of varying maximum b-value, number of shells, and Rician noise.

Main Results:

• QDTI measures, particularly α in white matter, showed b-value dependence, improving tissue contrast with higher b-values.
• 2-shell acquisitions introduced small systematic differences in QDTI measures compared to MbR.
• 3-4 shells at bmax=2000 s/mm² and 4 shells at bmax=3000 s/mm² minimized bias in D and α.

Conclusions:

• QDI enables robust parameterization of non-Gaussian diffusion in vivo.
• Optimized QDTI acquisition provides high reliability, accuracy, and precision within clinically feasible imaging times.
• This work facilitates advanced diffusion MRI analysis for clinical applications.