Not shooting an elephant

摘要

I am honored that Professors Gunn and Rabiner have seen fit to comment (PET neuroimaging: The elephant unpacks his trunk Comment on Cumming: "PET neuroimaging: The white elephant packs his trunk?") on my earlier commentary in the pages of Neuroimage (Cumming, 2014). Gored by my own tusk, this repartee can only properly begin with acknowledgment that they have detected an error in my figure, which significantly under-reported the historical publication rates for the search items "Brain AND PET". This error arose from the sloppy use of a spreadsheet, and may prove that real neurokineticists are not easily confounded by the logarithm. In any event, Gunn and Rabiner's Fig. 2 shows the true extent to which publication rates for brain fMRI have exceeded those for brain PET since the year 2002, currently by a factor of one unit of the natural logarithm in the literature as a whole; this factor is more like 20-fold in the recent pages of Neuroimage, as detailed in the main part of my commentary. Gunn and Rabiner's original consideration of PET publication rates by radionu-clide (Fig. 3) shows that the decline of Oxygen-15 studies is a driver for the falling rate of brain PET publications relative to fMRI and allied methods, and their Fig. 4 suggests a product-precursor relationship between the two methods. This latter observation is an elegant affirmation of my point that fMRI papers appearing in Neuroimage during 2012 frequently cited the precedent established by earlier PET studies of cerebral blood flow (as well as glucose and oxygen metabolism), which speaks for the maturation of the PET field.