Abstract
Point-resolved spectroscopy is commonly used for magnetic resonance spectroscopy (MRS) due to widespread clinical availability, yet its accuracy can be compromised by chemical shift displacement error (CSDE) and sensitivity to magnetic field inhomogeneities, particularly near cerebrospinal fluid (CSF)-rich areas. Semi-localization by adiabatic selective refocusing (sLASER) mitigates these issues through superior voxel localization, reducing susceptibility to CSDE and B1 inhomogeneity. However, direct clinical comparisons between PRESS and sLASER, especially under matched acquisition conditions, remain scarce. This study aimed to directly compare PRESS and sLASER sequences regarding metabolite quantification accuracy and spectral quality in brain regions adjacent to CSF, under identical water suppression scheme. Thirty healthy adult volunteers (17 men; mean age, 40 ± 9.6 years) were recruited between January and July 2023. MRS data were acquired with PRESS and sLASER sequences at the left medial thalamus (near the third ventricle), using identical voxel placement and water suppression scheme of VAPOR. Metabolite concentrations were quantified using LCModel, and spectral quality metrics including residual water peak height, spectral signal-to-noise ratio (SNR), and spectral linewidth were also analyzed. As a result, compared to PRESS, sLASER yielded significantly higher concentration of NAA + NAAG, whereas Gly + mI were significantly lower (all, FDR adjusted q < 0.05). Spectral SNR was significantly higher (+ 24%, P < 0.001), but residual water peak height and spectral linewidth did not differ significantly (P > 0.05). Variability analysis indicated significantly higher coefficient of variation for Glu + Gln with sLASER (P < 0.05). In conclusion, in CSF-adjacent region, sLASER significantly increased spectral SNR but also exhibited greater variability in concentration for specific J-coupled metabolite. Therefore, although sLASER improved certain spectral quality aspects, caution is warranted due to higher variability in metabolite quantifications.