[I003] Diffusion MRI: Sequence optimisation in oncology

Introduction Diffusion refers to the random Brownian molecular motion. In a bulk liquid, water molecules are free to move and this is characterized by a large value of the diffusion constant, D . The motion of water molecules within cells may be restricted by the cell wall and other structures, resulting in a lower value of D ; the same may be true of interstitial fluid in tissues if the cells are densely packed. Diffusion-Weighted Imaging (DWI) generates images based on the variation of D [1] . Diffusion-Weighted MRI DWI uses balanced pulses of magnetic field gradient, applied either side of a 180° pulse within a spin-echo pulse sequence, as introduced by Stejskal and Tanner in 1965 [2] . For stationary spins, any phase accrued during the first gradient pulse is unwound during the second pulse. Spins in random motion, however, become dephased, leading to a diffusion-dependent loss of signal. The signal can be given by S b , TE = M 0 exp - TE T 2 exp - bD where b is a factor which is dependent on the square of gradient strength ( G 2 ), its duration and spacing, known as the “ b -value”; the stronger the value of b , the higher is the diffusion weighting [3] . b values in the range 0–1000 s/mm 2 are typically used. In a DWI image with a large b -value, tissues containing freely-diffusing water molecules will appear dark (due to loss of signal), while tissues where diffusion is restricted will appear with greater intensity. Conclusions DWI is a standard part of the diagnostic armoury of clinical MRI, where one of its most important applications lies in the diagnosis of cancer. This lecture will describe the basic principles of DWI, with particular emphasis on its use in oncology [4] .