Prediction of the Degree of Metal Artifact using Localizer sequences of Magnetic Resonance Imaging in Lumbar spine

Motivation: Metal artifact reduction techniques needs longer scan time and shows incomplete MRI visibility for postoperative spinal evaluation. Goal(s): We assessed the reliability of localizer sequence image factors for predicting the metal artifact range in MRI. Approach: Multivariate ordinal logistic regression models were generated using prosthesis information and imaging factors measured on localizer images, to verify the factors correlated to the visibility of spinal canal and neural foramen on MRI. Results: Metal artifact degree at midline and foramen level of sagittal plane and inter-screw distance of coronal plane in localizer images were significant factors affecting metal artifact range of MRI. Impact: We found that the metal artifact degree in localizer sequence image can be correlated with the metal artifact range of diagnostic sequence image, and used to screen proper patients to apply metal artifact reduction sequence in postoperative spinal MRI.