Investigation of Background Phase Errors in Phase Contrast MRI

Abstract

Phase contrast MRI provides a accurate solution for evaluating blood ow. Unfortunately recent work has identi ed measurement errors due to the presence of background phase|a shading artefact that appear in MRI im- ages, caused mainly by induced eddy currents which distorts the magnetic eld and signi cantly impact on the clinical accuracy of ow assessments. The aim of this study was to develop a better understanding of background phase so that new methods of correction can be developed. Two studies were conducted. The rst study investigated the shape, linearity and tem- poral stability (spacial aspects) of background phase via phantom scanning at 1,296 di erent slice orientations using a standard phase contrast pulse sequence at minimum echo time (TE) and repetition time (TR). The second experiment investigated the temporal decay of background phase in uenced by induced eddy currents using a specially designed pulse sequence. Results showed that background phase is linearly shaped and is di erent for di erence slice orientations. Background phase is also highly dependent on the gradient waveforms applied to acquire the images. A rst order (pla- nar) surface t was able to approximate background phase to within 0.04 cms-1. A second order surface t was able to o er 50% better approxima- tion. Background phase is also stable over time. The median variation of background phase over an eight month period was 0.08 cms-1. Induced eddy currents responsible for background phase decay at long time constants lasting up to 3000 ms. Long time constants of decay were expected due to the low resistivity of the conducting materials inside the scanner at the cryogenic operating conditions of the scanner. Eddy currents were also observed to decay at multiple time constants and at opposing polarities giv- ing rise to bell-shaped temporal decay pro les. The temporal decay pro les of eddy currents however di ered for repetitions of the same experiment. This was believed to due to time-dependent parameters such as tempera- ture uctuations in the scanner due to resistive heating although further experimentation is needed to verify this.