Setting new tracks: not just creating another pretty picture

Diffusion tensor imaging (DTI) has the lure that functional MRI had in the 1980s; it often yields beautiful pictures, whether or not it confirms your beautiful hypotheses. It is not that we haven't learned much from both, indeed, functional MRI has revealed substantial insights into the networks of brain regions that work together to support cognitive functions; how these networks change to support recovery of function over time after stroke (Saur et al ., 2006); and how these networks are altered in disease states—for example, Alzheimer's disease, frontotemporal dementia (Zhou et al ., 2010) and developmental disabilities such as dyslexia (e.g. Hu et al ., 2010). Similarly, DTI, and particularly tractography, have provided important new information, for example, about the various connections of the superior longitudinal fasciculus, and how lesions affecting the separate connections might explain various types of language disorder (Catani and ffytche, 2005) or neglect (Bartolomeo et al ., 2007). It has also helped in monitoring disease progression in diseases such as multiple sclerosis; and proved useful in evaluating recovery in a variety of neurological contexts. But the lure to be avoided is to use DTI merely to localize a previously well described behaviour. That is, the field is already full of voxel-based morphometry studies of the precise voxels most associated with performance on a particular published test. Adding another layer of localization—the white matter tracts where disruption is associated with impaired performance on that particular test—is not a sufficient advance unless the behaviour itself is interesting, or a specific hypothesis about the white matter tract or the disease of interest is being tested.This issue of Brain remembers John Hughlings Jackson. In the 1860s Dr Hughlings Jackson wrote a large number of influential observations on …