Test-retest reliability of Diffusion Tensor Imaging metrics in neonates

Area of Science:

• Neuroimaging
• Developmental Neuroscience
• Medical Physics

Background:

• Diffusion tensor imaging (DTI) is crucial for studying brain microstructure but its application in neonates is limited due to myelination status and motion.
• Neonatal brains are largely unmyelinated, posing challenges for DTI reliability.

Purpose of the Study:

• To assess the intra-scan test-retest reliability of Diffusion tensor imaging (DTI) metrics in neonatal brain data.
• To evaluate the impact of analysis strategies on DTI metric reliability in infants.

Main Methods:

• A 96-direction DTI acquisition was divided into three segments for pairwise reliability analysis.
• Rigorous quality control was applied, selecting 25 diffusion encoding directions per segment from 86 infants.
• Tract-based spatial statistics (TBSS), voxelwise, and ROI analyses were employed.

Main Results:

• High intra-class correlation coefficient (ICC) values (ICC > 0.80) were observed for most DTI metrics.
• Residual motion, even after quality control, did not significantly associate with diffusion metrics.
• Analysis strategy and post-processing did not show differential effects on DTI metric reliability.

Conclusions:

• Diffusion tensor imaging (DTI) metrics are reliable in neonate brain data, even with low angular resolution common in infant scans.
• These findings support the use of neonatal DTI data in cross-sectional and longitudinal studies of brain white matter.
• Future research should focus on the reliability of fiber tracking techniques in neonatal DTI datasets.