MRI estimation of turbulent abnormalities in blood flow by improved multiparametric encoding gradients

A novel magnetic resonance approach to imaging turbulence in the velocity profiles in major blood vessels is presented. An nuclear magnetic resonance (NMR) image, in general, is dependent on parameters like r, v, a, which represent deterministic parameters of motion. In addition, spins undergo random motion such as diffusion and turbulence, which cannot be described by a predetermined set of parameters of motion. The differentiation between turbulent and nonturbulent flow is based on the deterministic or indeterministic nature of the flow as a function of time. A method of building up a general gradient from a basic set of gradients is investigated here. Image blurring due to the nondeterministic behavior of spins (e.g., diffusion and turbulence) is not removed by this technique of a composite gradient encoding field. This fact is the key to the possibility of identifying regions of turbulent flow. Because of its random nature, elimination of turbulent effects by use of a deterministic method is bound to fail. On the other hand, regions of turbulent flow may be identified owing to signal decrease from those regions.