Magnetic Resonance Imaging Hyperintensities in Alzheimer's Disease

To the Editor.  —This letter is a commentary on Leys and coworkers' article in the May 1990 issue of the Archives.¹ We commend Leys and coworkers on their effort to determine the significance of white matter and periventricular changes in patients with Alzheimer's disease. The pathological significance of these changes are particularly difficult to assess since, as Leys and coworkers indicate, they are found in many neurologically and psychiatrically normal controls. However, we feel that several methodological issues may make it difficult to generalize the findings in this study. First, the authors excluded 11 of 31 subjects due to poor quality in magnetic resonance imaging scans. They were not clear why more than one third of the patients with Alzheimer's disease had technical difficulties in imaging. We agree with the authors' concern that exclusion on the basis of an inability to cooperate with the magnetic resonance imaging scan may