Toward Precision and Reproducibility of Diffusion Tensor Imaging: A Multicenter Diffusion Phantom and Traveling...

Area of Science:

• Neuroimaging
• Biomarkers
• Precision Medicine

Background:

• Precision medicine utilizes quantitative biomarkers to minimize patient variability in disease diagnosis, treatment, and prevention.
• High-angular-resolution diffusion tensor imaging (DTI) is a key neuroimaging technique.

Purpose of the Study:

• To assess intersite variability of a harmonized 3T DTI protocol across 13 scanners from 11 academic medical centers.
• To evaluate the feasibility of standardizing DTI for the multi-site "Transforming Research and Clinical Knowledge in Traumatic Brain Injury" study.

Main Methods:

• Acquired diffusion MR imaging using a novel isotropic phantom and a traveling volunteer.
• Scanned data on thirteen 3T MRI scanners from GE Healthcare, Philips Healthcare, and Siemens.
• Calculated means and coefficients of variation for DTI parameters across scanners for each white matter tract.

Main Results:

• For the NIST phantom, the coefficient of variation for apparent diffusion coefficient was <3.8%.
• For the volunteer, coefficients of variation for primary DTI metrics averaged over the white matter skeleton were <5%.
• Large white matter tracts showed good reproducibility (<5% CV), while smaller tracts exhibited higher variability (up to 10% CV).

Conclusions:

• The study demonstrates the feasibility of standardizing DTI across different 3T MRI vendors.
• This standardization is essential for large-scale neuroimaging research studies.
• Harmonized DTI protocols can enhance the reliability of data for precision medicine in traumatic brain injury research.