As advocates of brain imaging, we believe its contribution to research and the clinic has already been significant. Yet, there has always been an ‘elephant in the room’when it comes to such studies in the pain field, particularly those using Functional Magnetic Resonance Imaging (fMRI) or event related electrophysiological recordings; they do not measure activity related to ongoing, background pain. Rather, they are techniques superbly capable of recording brain activity in response to an applied exogenous nociceptive stimulus or a provoked clinical symptom, often repeated and cycled between ‘‘pain on” and ‘‘pain off” periods. The brain pattern that emerges has been termed, for better but mostly for worse, the ‘pain matrix’, after Melzack’s neuromatrix concept was first usefully proposed [10]. Our problem with the ‘Pain Matrix’, as we have previously noted [15], is that people ‘pick ‘n’mix’what regions to include. Further, it gives the false impression there is something special about the matrix, that analogous to V1 for vision or S1 for touch, there is a ‘P1’for pain. Rather, it is a collection of brain regions. None are unique to pain, and many are involved in other aspects of perception and behaviour. Yet, when coordinated in activity the result is the multidimensional perception that is pain; a sensory, emotional, motivational and cognitive experience. Other brain regions may be recruited to exacerbate or reduce the dimensions of intensity and unpleasantness, providing a unique signature or profile for each individual [15] that we believe has translational, diagnostic relevance.

A possibly more interesting ‘matrix’to define, which might be also less subject to experimental bias, is one that reflects a major symptom in chronic pain states; namely, ongoing, background pain. fMRI studies, which use blood oxygen level dependant (BOLD) contrast as an indirect measure of changes in cerebral blood flow (CBF) coupled to neuronal activity [6], require a changing stimulus input, which effectively precludes its use to study tonic pain, although attempts have been made [1]. Positron Emission Tomography (PET)(H2 15O and 18FDG) is a technique amenable to examining blood flow and metabolic related responses to the tonic neural activity underpinning ongoing pain, however, it has been surprisingly limited in its application to this problem [2, 4, 5]. What we need is a noninvasive method, suitable for patients, that provides a quantitative measure of either blood flow or metabolic change that is linked to neuronal activity. That way, measures can be taken at different time periods when the ongoing pain is behaviourally different. Because these measures are quantitative they can be subtracted to provide the neural signature reflecting the change in ongoing pain. One such method exists, having been primarily developed for stroke applications: it is called arterial spin