Motivation: To improve the SNR efficiency of diffusion-prepared imaging by avoiding the use of a stabilizer gradient and its 2× signal penalty. Goal(s): Demonstrate that shot-to-shot motion-induced phases can be modelled as image-based amplitude variations in reconstruction. Approach: Monte Carlo simulations showed that eliminating the stabilizer gradient increases the number of excitations required to obtain accurate ADC estimates. Diffusion-prepared acquisitions with and without stabilizer gradients were performed in the brain and reconstructed with different retrospective NEX. Results: Eliminating the stabilizer gradient improved SNR at the cost of an increased minimum scan time, and diffusion contrast sensitivity to B1 inhomogeneity. Impact: Replacing the stabilizer gradient with use of a signal amplitude correction in diffusion-prepared imaging can reliably improve the SNR efficiency in cases where motion-induced and eddy-current-induced phase is controlled.
Lee et al. (Tue,) studied this question.