Accurate and sensitive measurements of magnetic susceptibility using echo planar imaging

Abstract

Susceptibility differences are common causes for artifacts in magnetic resonance (MR); therefore, it is important to choose phantom materials in a way that these artifacts are kept at a minimum. In this study, a previously proposed MR imaging (MRI) method [Beuf O, Briguet A, Lissac M, Davis R. Magnetic resonance imaging for the determination of magnetic susceptibility of materials. J Magn Reson 1996; Series B(112):111–118] was improved to facilitate sensitive in-house measurements of different phantom materials so that such artifacts can more easily be minimized. Using standard MRI protocols and distilled water as reference, we measured magnetic volume susceptibility differences with a clinical MR system. Two imaging techniques, echo planar imaging (EPI) and spin echo, were compared using liquid samples whose susceptibilities were verified by MR spectroscopy. The EPI sequence has a very narrow bandwidth in the phase-encoding direction, which gives an increased sensitivity to magnetic field inhomogeneities. All MRI measurements were evaluated in two ways: (1) manual image analysis and (2) model fitting. The narrow bandwidth of the EPI made it possible to detect very small susceptibility differences (equivalent susceptibility difference, Δχ_e≥0.02 ppm), and even plastics could be measured. Model fitting yielded high accuracy and high sensitivity and was less sensitive to other image artifacts as compared with manual image analysis.

Introduction

Susceptibility differences in an object can cause many problems in magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS). Near interfaces, the magnetic field is disturbed; therefore, the spatial encoding in MRI is affected. The signal displacements can appear as high-intensity spots and spots with signal void in the image due to accumulation or subtraction of signal. In MRS, a magnetic field change can lead to a shift in resonance frequency and peak broadening.

If it is possible to determine the susceptibility of phantom and implant materials, then it is also possible to minimize the artifacts in advance by choosing materials with small susceptibility differences. This can become even more important as the trend for clinical MR goes toward higher field strength, 3 T and beyond, and higher field strengths lead to more pronounced susceptibility effects.

By means of MRS, a susceptibility-induced shift in resonance frequency can be used to determine the volume susceptibility of signal-giving liquids [1]. The method is accurate and easy to use. However, phantoms are often made of plastics or other solid materials, which is why the use of the MRS method is not always applicable. Beuf et al. [2] devised an MRI method that can be used for any sort of material because it is not the object itself but the effect it has on a reference liquid that is imaged. The susceptibility difference between an outer compartment and an inner compartment in a coaxial circular cylindrical phantom was determined by the size of the artifact in an MR image. For ordinary imaging sequences, the sensitivity was limited and the susceptibility differences caused by materials usually used in phantoms (e.g., plastics) were not detectable.

The aims of this study were (1) to increase the sensitivity of the MRI measurements of susceptibility differences to make it possible to also measure plastics and tissue-like materials, (2) to improve the sensitivity and accuracy in data evaluation by designing an automatic evaluation program based on model fitting and (3) to apply the improved method to measure the volume susceptibilities of polymethyl methacrlyate (PMMA) and polyethylene.