Benchmarking the Impact of Anatomical Segmentation on In Vivo Magnetic Resonance Spectroscopy

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Quantitative MRI

Background:

• Accurate metabolite concentration estimation in brain magnetic resonance spectroscopy (MRS) relies on precise tissue segmentation.
• Variability in gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF) fraction estimates can arise from differences in segmentation software algorithms.
• This variability can propagate into metabolite concentration measurements, affecting their reliability.

Purpose of the Study:

• To investigate the impact of different brain segmentation software tools on the estimation of absolute metabolite concentrations in MRS.
• To quantify the variability introduced by segmentation methods in MRS data.
• To assess the influence of segmentation on biological benchmarks like age-related associations.

Main Methods:

• Three segmentation software tools (ANTs, FSL, SPM) were applied to an in vivo test-retest MR dataset.
• Differences in estimated tissue fractions (GM, WM, CSF) were analyzed.
• The propagation of these differences into tissue-corrected metabolite concentrations was evaluated.
• Age-related associations with GM and total creatine (tCr) were examined as a validity check.

Main Results:

• Significant differences (p < 0.0001) were found in tissue fraction estimates between the evaluated segmentation tools.
• These differences led to variations in metabolite concentration estimates of up to 9% under identical conditions.
• While the strength of correlation varied, no statistically significant differences were observed in age-related associations across methods.

Conclusions:

• The choice of segmentation methodology significantly contributes to the variability of absolute metabolite concentration estimates in MRS.
• Transparent reporting of segmentation methods is crucial for ensuring reproducibility and cross-study comparability in MRS research.
• Understanding segmentation-driven variability aids in distinguishing methodological artifacts from true biological effects.