Retrospective frequency drift correction of rosette MRSI data using spectral registration

Area of Science:

• Magnetic Resonance Imaging
• Spectroscopy
• Medical Physics

Background:

• Frequency drift correction is crucial for Magnetic Resonance Spectroscopy (MRS) and Spectroscopic Imaging (MRSI).
• Traditional MRSI drift correction requires separate navigator scans, complicating the process due to phase-encoding gradients.
• Developing navigator-free drift correction methods is essential for efficient MRSI data processing.

Purpose of the Study:

• To demonstrate a novel self-navigating rosette MRSI technique for retrospective frequency drift correction.
• To eliminate the need for separately acquired navigator echoes in MRSI.
• To improve spectral quality and metabolite quantification in vivo MRSI data.

Main Methods:

• Implementation of a self-navigating rosette MRSI sequence for brain imaging in healthy volunteers.
• Isolation of center k-space FIDs from each rosette acquisition shot.
• Application of time-domain spectral registration for frequency offset estimation and retrospective correction.

Main Results:

• Significant improvements in signal-to-noise ratio (12.9%) and spectral linewidths (18.5%) after drift correction.
• Reduction in average Cramer-Rao lower bounds uncertainty estimates by 5.0% for all metabolites.
• Demonstrated feasibility of retrospective correction without navigator echoes.

Conclusions:

• Self-navigating rosette MRSI combined with time-domain spectral registration enables effective retrospective frequency drift correction.
• This approach significantly enhances spectral quality and metabolite quantification in in vivo MRSI.
• The developed method offers a more efficient alternative to traditional navigator-based drift correction techniques.