Minimizing artifacts caused by metallic implants at MR imaging: experimental and clinical studies.

Abstract

OBJECTIVE: The purpose of this study was to investigate the effect of metallic implant positioning on MR imaging artifacts, to determine the optimal imaging conditions for minimizing artifacts, and to show the usefulness of artifact-minimizing methods in imaging of the knee. MATERIALS AND METHODS: Using MR images of experimental phantoms (titanium alloy and stainless steel screws), we compared the magnitude of metal-induced artifacts for various pulse sequences, different imaging parameters for the fast spin-echo sequence, and different imaging parameters for several incremental angles between the long axis of the screw and the direction of the main magnetic field. In clinical MR imaging of knees with metallic implants (n = 19), we assessed geometric distortion of anatomic structures to compare the influence of different pulse sequences (n = 19), frequency-encoding directions (n = 7), and knee positions (n = 15). RESULTS: Titanium alloy screws consistently produced smaller artifacts than did stainless steel screws. In experimental MR studies, artifacts were reduced with fast spin-echo sequences, with a screw orientation as closely parallel to the main magnetic field as possible, and, particularly, with smaller voxels that correlated positively with artifact size (R2 = .88, p < .01). In clinical MR studies, fast spin-echo MR imaging obscured articular structures less than did spin-echo imaging (8/19 patients). In particular, the anterior-posterior frequency-encoding direction (3/7 patients) and the flexion position of the knee (5/15 patients) were effective in reducing artifacts. CONCLUSION: MR artifacts can be minimized by optimally positioning in the magnet subjects with metallic implants and by choosing fast spin-echo sequences with an anterior-posterior frequency-encoding direction and the smallest voxel size.