Variability of multidimensional diffusion-relaxation MRI estimates in the human brain.

Area of Science:

• Biomedical Imaging
• Neuroimaging
• Magnetic Resonance Imaging

Background:

• Conventional MRI techniques average diffusion and relaxation properties within voxels, potentially masking subtle cellular changes.
• Diffusion tensor imaging (DTI) and relaxometry provide valuable but limited information about tissue microstructure.

Purpose of the Study:

• To characterize whole-brain multidimensional MRI (MD-MRI) distributions and derived parameters.
• To evaluate the intrascanner test-retest reliability, repeatability, and reproducibility of MD-MRI.
• To explore the potential of MD-MRI as a neuroimaging biomarker for enhanced sensitivity and specificity.

Main Methods:

• MD-MRI was employed to replace voxel-averaged diffusion tensor quantities and R1/R2 relaxation rates with multidimensional distributions.
• Analysis focused on selective extraction and mapping of specific diffusion-relaxation spectral ranges corresponding to cellular features.
• Comparison of white matter tracts and cortical/subcortical gray matter regions was performed.

Main Results:

• Notable variations in diffusion-relaxation profiles were observed between white matter, gray matter, and cerebrospinal fluid, indicating unique microscopic characteristics.
• MD-MRI-derived diffusion and relaxation mean parameters demonstrated reliability and repeatability comparable to conventional DTI and relaxometry.
• Estimated signal fractions of intravoxel spectral components were reproducible, supporting the viability of spectral analysis for MD-MRI data.

Conclusions:

• A clinically feasible MD-MRI protocol reliably captures rich structural and chemical information within voxels.
• MD-MRI shows significant potential for developing novel MRI biomarkers with improved sensitivity and specificity.
• The spectral analysis approach to MD-MRI data is viable for clinical neuroimaging applications.