Improved Susceptibility Weighted Imaging methodusing multi-echo acquisition

Abstract

Susceptibility weighted imaging (SWI) emphasizes the magnetic susceptibility differences between brain tissues. However, the method is also sensitive to non-local susceptibility effects from background field inhomogeneity, particularly in areas near air-tissue interfaces. These large-scale effects often manifest as signal artifacts in areas such as the frontal and temporal lobes. In this article, new acquisition and post-processing approaches are proposed to remove background field inhomogeneity artifacts in both magnitude and phase data for improved SWI coverage. The proposed method acquires three echoes in a 3D gradient echo (GRE) sequence, with a field compensation gradient (z-shim gradient) applied to the third echo. The artifacts in the magnitude data are compensated by signal estimation from all three echoes. The artifacts in phase signals are removed by modeling the background phase distortions using Gaussians. The method was applied in vivo and compared with conventional SWI. The results demonstrate that this method successfully compensates for background field inhomogeneity artifacts in magnitude and phase images, and improves the quality of SWI images. In particular, vessels in frontal lobe, which were not observed in conventional SWI, were identified in the proposed method.