Retrospective motion compensation for edited MR spectroscopic imaging.

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Spectroscopy

Background:

• Edited magnetic resonance spectroscopic imaging (MRSI) enables mapping of low-concentration metabolites like GABA and GSH.
• Head motion and instabilities during MRSI acquisition can cause significant subtraction artifacts, compromising data integrity.

Purpose of the Study:

• To develop and validate a retrospective motion compensation algorithm for edited MRSI.
• To significantly reduce subtraction artifacts in MRSI data caused by motion.

Main Methods:

• A novel algorithm identifies motion-affected MRSI signals by analyzing residual water and lipid peaks in k-space.
• Affected spectra are corrected (phase correction, replacement, or removal) before spatial Fourier transformation.
• The algorithm was tested on GABA-edited spin-echo MRSI data from 8 healthy adults at 3T and a GSH-edited MRSI acquisition.

Main Results:

• Motion-compensated MRSI data showed significantly fewer subtraction artifacts compared to non-compensated data.
• The normalized choline (Cho) peak area, an indicator of artifact severity, was 5.14 times lower with motion compensation.
• A 'removal-only' approach also demonstrated effectiveness for motion-corrupted GSH-edited MRSI data.

Conclusions:

• Retrospective motion compensation in k-space is feasible for edited MRSI.
• The proposed algorithm effectively reduces subtraction artifacts, enhancing the quality of metabolite quantification in MRSI.