Multi-parametric neuroimaging reproducibility: a 3-T resource study.

Area of Science:

• Neuroimaging
• Radiology
• Medical Physics

Background:

• Modern MRI enables detailed brain analysis, but multi-parametric studies face reproducibility challenges due to limited data.
• Combining different MRI protocols (e.g., functional, anatomical, connectivity) is a growing trend.

Purpose of the Study:

• To develop and validate a clinically feasible 60-minute MRI protocol for comprehensive, multi-parametric human brain assessment.
• To provide scan-rescan reproducibility data for various brain regions and imaging contrasts.

Main Methods:

• A novel 60-minute, 3-T MRI protocol was optimized for quantitative, morphometric, functional, and micro-architectural brain assessments.
• Scan-rescan reproducibility was evaluated in 21 healthy volunteers across multiple brain structures.
• Region of interest (ROI) based analysis was used to quantify mean intensity, variability, and reproducibility for each contrast.

Main Results:

• The protocol achieved a mean volume-wise reproducibility of 3.5% across key brain structures like the cortex, white matter, and brainstem.
• Structural imaging showed high consistency (~1-5% variability), while diffusion and other quantitative scans had higher variation (~<10%).
• Specific sequences showed higher variability in certain structures (e.g., ASL in white matter, quantitative T2 in caudate), partly due to automated ROI placement.

Conclusions:

• The developed multi-parametric MRI protocol offers a reproducible and comprehensive approach to brain imaging within a clinically feasible timeframe.
• Publicly available data and analysis routines will facilitate algorithm optimization and the development of advanced neuroimaging techniques.
• This work establishes a crucial baseline for future multi-parametric imaging protocol development and validation in neuroscience research.