[Functional imaging of pain].

In this review, we summarize the contribution of functional imaging to the question of nociception in humans. In the beginning of the 90's, brain areas supposed to be involved in physiological pain processes essentially concerned the primary somatosensory area (SI), thalamus, and anterior cingulate cortex. In spite of these a priori hypotheses, the first imaging studies revealed that the main brain areas and those providing the most consistent activations in pain conditions were the insular and the SII cortices, bilaterally. This has been checked with other techniques such as intracerebral recordings of evoked potentials after nociceptive stimulations with laser showing a consistent response in the operculo-insular area whose amplitude correlates with pain intensity. In spite of electrode implantations in other areas of the brain, only rare and inconsistent responses have been found outside the operculo-insular cortices. With electrical stimulation delivered directly in the brain, it has also been shown that stimulation in this area only - and not in other brain areas - was able to elicit a painful sensation. Thus, over the last 15 years, the operculo-insular cortex has been re-discovered as a main area of pain integration, mainly in its sensory and intensity aspects. In neuropathic pain also, these areas have been demonstrated as being abnormally recruited, bilaterally, in response to innocuous stimuli. These results suggest that plastic changes may occur in brain areas that were pre-defined for generating pain sensations. Conversely, when the brain activations concomitant to pain relief were taken in account, a large number of studies pointed out medial prefrontal and rostral cingulate areas as being associated with pain controls. Interestingly, these activations may correlate with the magnitude of pain relief, with the activation of the peri-acqueductal grey (PAG) and, at least in some instances, with the involvement of endogenous opioids.