Effects of different field strengths, gradient directions, and acquisitions on fractional anisotropy in diffusion tensor imaging: A tract-based spatial statistics study

The effects of increasing acquisitions and gradient directions on fractional anisotropy (FA) at 1.5T versus 3T were investigated using an operator-independent methodology (tract-based spatial statistics) by minimizing technical differences in image acquisition between the two magnetic resonance imaging platforms and also FA values at roughly equivalent signal-to-noise ratio and data acquisition time were compared. Five healthy subjects underwent this experiment. Diffusion tensor imaging (DTI) was performed at 1.5T and 3T. Gradient schemes were 6, 30, and 96 directions. The numbers of acquisitions were 5, 10, and 15 for 6-direction; 1, 2, and 3 acquisitions for 30-direction; and 1 acquisition for 96-direction scheme. FA values in the callosal genu (CG), callosal splenium (CS), posterior limb of the internal capsule (PLIC), and the superior longitudinal fasciculus (SLF) bilaterally were assessed. For 6-direction scheme at 5 acquisitions, FA values were higher at 1.5T than at 3T. FA values derived using 30-direction scheme were higher at 3T than 1.5T in the left SLF and PLIC bilaterally. In the CG and CS, FA values at equivalent signal-to-noise ratio derived from different acquisitions and gradient directions differed by <7.6%. In the right PLIC and left SLF, FA values measured for 6-direction scheme differed from 30-direction scheme by >11% at 1.5T. DTI data acquired at 3T are not always comparable to that at 1.5T. The center of the brain does not appear to be affected but the peripheral white matter appears to be affected by field strengths, number of acquisitions, and gradient directions. (c) 2013 Wiley Periodicals, Inc. Concepts Magn Reson Part B (Magn Reson Engineering) 43B:4148, 2013