Magnetic Resonance Imaging of the Central Nervous System

Abstract

To the Editor.— The American Medical Association's Council on Scientific Affairs 1 sings the praises of magnetic resonance imaging (MR I) and indeed this praise is well deserved in many areas. Imaging of the posterior fossa and intrinsic spinal cord lesions is clearly superior with MR I. However, the enthusiasm for this new tool needs to be tempered in at least two very important areas. In the evaluation of multiple sclerosis, MRI is much more sensitive in detecting abnormalities than computed tomography, but an abnormal study result is not synonymous with the disease. The diagnosis remains a clinical one. 2 In the study by Ormerod et al 3 of disseminated lesions in isolated optic neuritis, three of 37 controls had lesions on MRI consistent with multiple sclerosis. Kent and Larson 4 estimated a false-positive rate of 0.06 (false-positive studies divided by all positive studies in the studied population) from the