TO THE EDITOR: I have read with great interest and enthusiasm the recent article by Chen et al. on the rapid scanning protocol for brain 18F-FDG PET (1). They found that rapid scanning for brain 18F-FDG PET is nearly equivalent to conventional scanning in the diagnosis of Alzheimer's disease. In their study, rapid scanning was performed later than conventional scanning after 18F-FDG injection, indicating the advantage of late scanning. Because the receiver-operating characteristic curves indicated that the area under the curve of rapid (late) scanning was slightly larger than that of conventional (early) scanning, their report supports our previous finding that late scanning is superior to early scanning in detecting hypometabolic regions in patients with Alzheimer's disease (2). I am also pleased that their study found regional 18F-FDG uptake differences between early and late scanning, verifying findings previously reported by my group (3). That is, relative 18F-FDG uptake in the posterior cingulate and parietal cortices, which are the regions affected in the Alzheimer's diseased brain, are larger at late scanning than at early scanning and 18F-FDG uptake in the cerebellum is lower at late scanning than at early scanning. We considered that these differences between the 2 sets of scans might have been related to regional differences in rate constants, such as K1, which indicates 18F-FDG transportation from plasma to tissue, and k3, which indicates phosphorylation of 18F-FDG (2). I expect this interesting brain physiology to be investigated further by using parametric mapping with compartment models.