Back

Absolute quantification of cerebral metabolites using 2D 1H-MRSI with quantitative MRI-based water reference

Thomas, D. C.; Alcicek, S.; Manzhurtzev, A.; Hattingen, E.; Wenger, K. J.; Pilatus, U.

2024-11-25 neuroscience
10.1101/2024.11.25.625224 bioRxiv
Show abstract

PurposeMetabolite concentrations are valuable biomarkers in brain tumors (BT). However, correction of water relaxation effects often requires time-consuming quantitative MRI (qMRI) sequences on top of a lengthy spectroscopic water reference acquisition. The goal of this work was to develop and validate a fast metabolite quantification method where a 2D spectroscopic water reference acquisition is obtained using a fast qMRI protocol and single-voxel STEAM sequence. MethodsA 2D sLASER sequence was acquired for MRSI. An 8-minute qMRI protocol was also acquired. A single-voxel unsuppressed water signal was acquired using a STEAM sequence. The H2O map, obtained from qMRI, was calibrated based on the STEAM-signal to obtain the spectroscopic water reference (proposed method). Five healthy volunteers and one BT patient were scanned at 3T. Concentrations obtained using the proposed and two reference methods, one where water relaxation effects were corrected using literature values (Reference method) and one where they were corrected using qMRI-derived values (Reference method with qMRI) were compared. ResultsIn healthy subjects, WM metabolite concentrations obtained using water relaxation using literature values (Reference method) significantly differed from those using individual-specific corrections (Reference method with qMRI and proposed method). Bland-Altman analyses revealed a very low bias and SD of the differences between the Reference method with qMRI and the proposed-method (Bias<0.5% and SD<10%). The BT regions showed a [~]15% underestimation of metabolite concentrations using the Reference method. ConclusionFor metabolite quantification, accurate water referencing with individual-specific corrections for water relaxation times was obtained in 8 minutes using the proposed method.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.