Reproducibility of quantitative magnetization-transfer imaging parameters from repeated measurements

Magn Reson Med. 2010 Aug;64(2):391-400. doi: 10.1002/mrm.22350.


Quantitative magnetization-transfer imaging methods provide in vivo estimates of parameters of the two-pool model for magnetization-transfer in tissue. The goal of this study was to evaluate the reproducibility of quantitative magnetization-transfer imaging parameter estimates in healthy subjects. Magnetization-transfer-weighted and T(1) relaxometry data were acquired in five healthy subjects at multiple time points, and the variability of the resulting fitted magnetization-transfer parameters was evaluated. The impact of subsampling the magnetization-transfer data and correcting field inhomogeneities was also evaluated. The key parameters measured in this study had an average variability, across time points, of 4.7% for the relative size of the restricted pool (F), 7.3% for the forward exchange constant (k(f)), 1.9% for the free pool spin-lattice relaxation constant (R(1f)), 4.5% for the T(2) of the free pool (T(2f)), and 2.3% for the T(2) of the restricted pool (T(2r)). Our findings show that serial quantitative magnetization-transfer imaging experiments can be performed reliably, with good reproducibility of the model parameter estimates, and demonstrate the reproducibility of acquisition schemes with fewer magnetization-transfer contrasts. This establishes the feasibility of this technique for monitoring patients affected by degenerative white matter diseases while providing critical data to estimate the statistical power of such studies.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Algorithms*
  • Brain / pathology*
  • Female
  • Humans
  • Image Enhancement / methods
  • Image Interpretation, Computer-Assisted / methods*
  • Information Storage and Retrieval / methods*
  • Male
  • Multiple Sclerosis / pathology*
  • Reproducibility of Results
  • Sensitivity and Specificity
  • Young Adult