Ralph Sinkus1, Simon Auguste Lambert1, Lucas Hadjilucas1, Shaihan Malik2, Anirban Biswas1, Francesco Padormo2, Jack Lee1, and Joseph V Hajnal2
1Imaging Sciences & Biomedical Engineering Division Kings College, King's College London, London, United Kingdom, 2Centre for the Developing Brain & Department Biomedical Engineering, King's College London, London, United Kingdom
Classical DWI methods extract information about
microstructural tissue complexity from the signal decrease of the MR-magnitude
as a function of b-value. Utilization of linear gradients for motion encoding
prevents theoretically the use of the MR-phase. Rather, the diffusion
information is encoded in the MR-magnitude via global spin dephasing due to
Brownian motion with zero net phase shift. This dogma is overturned when
considering quadratic gradient fields in space. We demonstrate in theory,
experiment, and simulation that the diffusion process leads to a net phase
shift with minimal loss in signal magnitude when imaging at the minimum of the
quadratic gradient.
