Material Orientation Artifact Studies in Magnetic Resonance Imaging

With the increased interest of MRI guided interventional procedures in modern medical treatments, image distortion and artifact formation based on material selection and orientation within the MRI scanner are central concerns for precise object localization. The goal of this study was to illustrate the artifact behavior of materials with various magnetic susceptibilities and radio frequency conductivity values corresponding to object orientation relative to the primary magnetic field. To test the effects of orientation on image distortion and image artifacts, 0.125 inch cylindrical test samples of various materials were imaged using a clinical Siemens 3 Tesla MR scanner. Modern medical instrumentation and surgical utensils are typically made from highly paramagnetic materials (e.g., titanium, nitinol, or stainless steel) which also have high RF conductivities. The combination of these two material properties cause both primary magnetic field (B0) and RF field (B1) inhomogeneities which lead to local image distortions. A change in the local magnetic field induces errors within the slice selection gradient, as the precessional frequency of the proton nuclei in the desired region of interest will not correspond to the exact spatial location on the object and will excite a broader region due to the RF conductivity of the material. Conversely to more traditional surgical materials, diamagnetic materials (e.g., bismuth, pyrolytic carbon, water, most plastics) are free from the susceptibility artifacts due to B0 inhomogeneities and thus offer a level of MR compatibility that traditional materials cannot. A specific testing phantom was built to fit a clinical wrist coil. The phantom consisted of an aqueous solution of gadolinium and copper sulfate to increase image contrast and a rotatable turret post for sample positioning. The particular materials studied were chosen to demonstrate the wide variation in both magnetic susceptibility values and RF conductivities (e.g., 6A1-4V titanium, 316L stainless steel, carbon fiber, 6061 T6 aluminum, brass, copper, beryllium copper). ImageJ software measured the overall pixel area and major dimension of each MR image artifact at 0, 45, and 90 degree orientations of each test sample relative to B0. The results of the measurements indicated measurable increases in signal are of the paramagnetic and highly conductive test specimens orientated orthogonal to the primary magnetic field. For instance, two common medical grade materials such as 316L stainless steel and 6Al-4V titanium resulted in artifact area increases of 770±10% and 234±10%, respectively, relative to the actual cross sectional area of the sample. Conversely, the more diamagnetic materials, carbon fiber and beryllium copper demonstrated increased artifact areas of 8±10% and 12±10%, respectively. Errors in artifact area percentage growth measurement are primarily attributed to manual image segmentation and variation in coil positioning within the MRI bore. The results indicate that MR image artifact size and object distortion characteristics can be influenced by both material selection and object orientation relative to the primary magnetic field. In the interest of accurate navigation of image guided equipment and devices, interventional devices should be tested for image distortion in multiple orientations. This work is supported by MIMTeC, a National Science Foundation Industry University Collaborative Research Center and by NIH Grant P30 NS057091.