0.6 T Magnetic Resonance Imaging of the Orbit

Abstract

Magnetic resonance (MR) imaging of the orbit was performed with a 0.6 T superconducting imaging system in 100 patients with normal orbits who were being evaluated for brain pathology and in 21 additional patients with a variety of orbital lesions to determine the efficacy of MR imaging in displaying orbital abnormalities. Usually, MR studies were performed using a multislice technique with multiple spin-echo pulse sequences and 30, 60, and 90 msec echo times and 500, 1500, and 2000 msec repetition times. Using section thicknesses of about 8 mm, imaging was performed in the transaxial, coronal, and sagittal projections. Pixel size was 0.9 × 1.8 mm, and the examination took about 30 min. The MR findings were compared with computed tomographic (CT) findings in all cases. Either combined axial and coronal studies of 5-mm-thick sections or a thin axial study of 1-mm-thick sections followed by reformatting techniques to obtain multiplanar images was used. Contrast enhancement was used in the CT studies. Both MR and CT clearly demonstrated the soft-tissue abnormality in all cases except two, in which MR failed to detect the abnormality. In one, MR failed to detect a small retrobulbar hemorrhage that occurred after a surgical procedure for retinal detachment. In the second case, rather extensive calcification in the posterior choroidal layers and lens was not detected by MR imaging. In several other cases, MR provided information beyond that obtained with CT. MR has the advantage of providing exquisite anatomic detail in multiplanar images, and it appears to be more sensitive than CT in detecting small, subacute and chronic hemorrhage within soft-tissue masses in the orbit and in detecting ischemia of the globe. CT is superior to MR imaging in portraying fine bone detail.