A framework for optimization-based design of motion encoding in magnetic resonance elastography.

PurposeIn conventional three-dimensional magnetic resonance elastography, motion encoding gradients (MEGs) synchronized to a mechanical excitation are applied separately in each direction to encode tissue displacement generated by the corresponding waves. This requires long acquisition times that introduce errors due to patient motion and may hinder clinical deployment of magnetic resonance elastography. In this article, a framework for MEGs sequence design is proposed to reduce scanning time and increase signal-to-noise ratio. Theory and MethodsThe approach is based on applying MEGs in all three directions simultaneously with varying parameters, and formulation of the problem as a linear estimation of the wave properties. Multidirectional MEGs sequences are derived by setting the problem in an experimental design framework. Such designs are implemented and evaluated on simulation and phantom data. ResultsEstimation error of the displacement using the proposed MEGs designs is reduced up to a factor of two in comparison with a unidirectional design with a same number of acquisitions. Alternatively, for the same error, scanning time is reduced up to a factor of three using the multidirectional designs. ConclusionThe proposed framework generalizes acquisition of magnetic resonance elastography, and allows quantification of design performance, and optimization-based derivation of designs. Magn Reson Med 73:1514-1525, 2015. (c) 2014 Wiley Periodicals, Inc.