Test-retest reliability and concurrent validity of in vivo myelin content indices: Myelin water fraction and calibrated T1 w/T2 w image ratio

Abstract

In an age-heterogeneous sample of healthy adults, we examined test-retest reliability (with and without participant repositioning) of two popular MRI methods of estimating myelin content: modeling the short spin-spin (T₂ ) relaxation component of multi-echo imaging data and computing the ratio of T₁ -weighted and T₂ -weighted images (T₁ w/T₂ w). Taking the myelin water fraction (MWF) index of myelin content derived from the multi-component T₂ relaxation data as a standard, we evaluate the concurrent and differential validity of T₁ w/T₂ w ratio images. The results revealed high reliability of MWF and T₁ w/T₂ w ratio. However, we found significant correlations of low to moderate magnitude between MWF and the T₁ w/T₂ w ratio in only two of six examined regions of the cerebral white matter. Notably, significant correlations of the same or greater magnitude were observed for T₁ w/T₂ w ratio and the intermediate T₂ relaxation time constant, which is believed to reflect differences in the mobility of water between the intracellular and extracellular compartments. We conclude that although both methods are highly reliable and thus well-suited for longitudinal studies, T₁ w/T₂ w ratio has low criterion validity and may be not an optimal index of subcortical myelin content. Hum Brain Mapp 38:1780-1790, 2017. © 2017 Wiley Periodicals, Inc.

Keywords: MRI; Simpson paradox; T2 relaxation; brain water mobility; white matter.

Figure 1

Design diagram of the reliability trials.

Figure 2

Plots of associations between Myelin Water Fraction, geomT_2iew and T₁w/T₂w ratio (multiplied by 100) for all regions of interest, and their aggregation. Values are based on the mean of the right and left side except for the genu and splenium. *P < 0.05; **P < 0.008. [Color figure can be viewed at http://wileyonlinelibrary.com]