Human In Vivo Validation of Frequency-Dependent QTI

Area of Science:

• Neuroimaging
• Biophysics
• Medical Physics

Background:

• Quantitative MRI techniques like q-space trajectory imaging (QTI) offer insights into tissue microstructure.
• Accurate estimation of QTI metrics often relies on approximations, such as the centroid frequency approximation, which may not hold under realistic conditions.

Purpose of the Study:

• To investigate the design of QTI waveforms for probing QTI metrics at a single centroid frequency.
• To assess the accuracy of this approximation under realistic in vivo human brain imaging conditions.

Main Methods:

• Designed realistic diffusion encoding waveforms using double-rotation gradient and magic-angle spinning techniques.
• Varied encoding bandwidth, tuning (alignment of encoding spectra across encodings), and spectral isotropy (alignment across axes).
• Computed QTI metrics using analytical diffusion spectra and validated waveforms with a multiband spiral spin-echo sequence in vivo.

Main Results:

• Simulations revealed frequency-dependent QTI metrics.
• Combining tuning and spectral isotropy yielded metrics close to the ground-truth at the centroid frequency.
• Realistic waveforms with broader bandwidth introduced minimal additional error in QTI metrics.
• In vivo measurements qualitatively confirmed simulation findings regarding spectral properties.

Conclusions:

• The accuracy of the centroid frequency approximation is dependent on encoding spectra and tissue properties.
• Employing similar encoding spectra for tuning and spectral isotropy enables the measurement of QTI metrics characteristic of the centroid frequency.