How to perform diffusion-weighted imaging

Diffusion-weighted imaging (DWI) has become an invaluable tool in the management of patients with stroke. DWI relies on detecting the random diffusion of water molecules. In normal tissues this movement may be restricted by the presence of cellular structures, which provide a barrier to free movement. This occurs in myelinated white matter, where movement is restricted more across than along fibres. This directional dependence is termed anisotropic restricted diffusion. The diffusion of water molecules can be made the dominant contrast mechanism within an image by applying large magnetic field gradients. The pulsed gradient spin echo (PGSE) sequence provides sensitivity to diffusion with gradient pulses either side of the 180 degrees refocusing pulse. This sequence is generally heavily T2 weighted. In order to detect normal anisotropic properties within the different components of the medullary core, association, commissural and projection fibres, DWI must be performed with sensitisation in at least three directions. PGSE sequences have been used to obtain the diffusion coefficient (D*), a measure of mobility at the molecular level within tissue. In acute infarction D* is increased; in brain death it is decreased. Diffusion contrast needs to be optimised in relation to the highly T2-dependent nature of PGSE sequences. This also requires a more detailed knowledge of how D* changes in disease, but information on nonischaemic neurological conditions is still very limited.