Uncertainty estimations for quantitative in vivo MRI T1 mapping

Area of Science:

• Neuroimaging
• Quantitative MRI
• Biomedical Engineering

Background:

• Longitudinal relaxation time (T(1)) mapping of brain tissue is crucial for clinical research and MRI sequence development.
• Accurate interpretation of in vivo T(1) variations requires understanding the uncertainty associated with quantitative T(1) parameters.
• Current methods may not fully capture the inherent variability in T(1) measurements.

Purpose of the Study:

• To present a general framework for estimating uncertainty in quantitative T(1) mapping.
• To validate a novel approach combining a slice-shifted multi-slice inversion recovery EPI technique with the statistical wild-bootstrap method.
• To evaluate T(1) uncertainty in specific brain regions of healthy volunteers.

Main Methods:

• Utilized a slice-shifted multi-slice inversion recovery EPI technique.
• Implemented the statistical wild-bootstrap approach for uncertainty estimation.
• Performed simulations and experimental analyses on four healthy volunteers.

Main Results:

• The proposed framework successfully estimated T(1) uncertainty without requiring repeated measurements.
• Variation in T(1) within anatomical regions of similar tissue types exceeded measurement uncertainty.
• Tissue heterogeneity and partial volume effects were identified as primary drivers of observed T(1) variability.

Conclusions:

• The developed method provides a reliable estimation of T(1) uncertainty in brain MRI.
• Observed T(1) variability is largely attributable to biological tissue characteristics rather than measurement limitations.
• This approach can aid in calculating significant effect sizes for group comparisons in future studies.