MR Flow Imaging A Study of Flow Effects on the Gradient Echo Sequence

Effects of flow are a fundamental part of magnetic resonance imaging (MRI). Since almost every part of the body contains some kind of flow, the effects are seen in generally all MR images. The presence of flow in an image slice may be a vast source of artefacts but the effects of flow can also be used for diagnostic purposes, for example in magnetic resonance angiography (MRA). In MRA, flow effects are utilised in order to obtain vascular information from a patient. MRA has become aneffective diagnostic tool capable of obtaining high resolution morphological and physiological vascular information throughout the body. The aim of this thesis has been to study the origin of flow effects in MRI, and to investigate the behaviour of MR signal when flow is present. Of special interest was effects in the spoiled gradient echo pulse sequence, since it is the most commonsequence used in MRA. The signal equations for the spoiled gradient echo sequence were used to study signal behaviour when flow is present and sequence and tissue parameters are changed. Calculations were performed assuming both stationary and flowing tissue. A program for MS Windows 95™ called GEsignal was developed especially for this work. It was written in a visual object oriented program language (Borland Delphi). GEsignal calculates the signal of the spoiled gradient echo sequence. It can be used for both stationary tissue and for flow perpendicular to an image slice. Phase effects on the magnetisation vector due to flow were also studied. A spreadsheet (MS Excel) was set up to illustrate these effects. The spreadsheet can be used to vary up to four gradient lobes in time and amplitude strength. It is updated in real time to show phase effects due to flow in the direction of the gradients. Simulations and flow measurements were performed to verify the calculations. The Simulations were carried out using “The MRI Simulator”, a PC based program that simulates a real MR scanner (Petersson et al, 1993). The flow measurements were carried out at a Siemens Magnetom Vision 1.5T MR scanner. Calculations, simulations and measurements agreed fairly well qualitatively. Quantitatively, calculations and simulations agreed well. The signal difference was a few percent except for the combination of high flip angles and high flow velocity, were the difference may be as high as 10 to 15%. Measured data differed rather much from calculated results. However, if the calculated signal was modified to take into account a gaussian flip angle variation over the image slice, the results agreed very well with measured signal. The signal difference between calculations and measurements was reduced from tens of percent to a few percent by introducing a gaussian flip angle distribution. Both GEsignal and the spreadsheet are useful tools to study flow effects qualitatively. If a quantitative study is to be performed, the method used in GEsignal appears too restricted. However, the result may be modified to agree well with measured data. The spreadsheet finds its best application as a tool to understand the concept of gradient moment nulling