Omitted Author Affiliation

Tim V. Salomons,Gian Domenico Iannetti,Meng Liang

DOI: https://doi.org/10.1001/jamaneurol.2016.3119

IF: 29.907

2016-01-01

JAMA Neurology

Abstract:concerns would be critical.As no such claims were made, however, these concerns largely miss the point of our letter.We concluded that pain matrix activation is insufficient evidence for the presence of pain.This conclusion does not hinge on the null finding (lack of group differences).Rather, it requires demonstrating robust pain matrix activation in the absence of pain.Patients with loss-of-function SC9A mutations are extremely rare (limiting sample size), but studying this population allows us to conclusively rule out pain as an explanation of the measured neural response.Testing for significant differences and displaying group scatter plots merely allows us to demonstrate that patient responses are within the same range as those observed in individuals who experienced pain in response to an identical stimulus.To strengthen inferences about nonspecificity, we include an analysis based on a reverse (rather than forward) inference mask of pain (Figure).Many in the neuroimaging field feel the nonspecificity of the pain matrix has already been conclusively demonstrated 2 and widely accepted.We wish this were true, but recent scientific debate over the "selectivity" of subsets of the pain matrix, 3 controversy over the use of neuroimaging as medicolegal evidence of pain, 4 and popular media reports conflating pain matrix activation with the experience of pain 5 demonstrate that pain matrix activation continues to be used as evidence for pain, both in the scientific community and in the court of public opinion.Finally, we disagree that parametric designs are a remedy for unspecific confounds.We have used these powerful designs in our own work to isolate responses that track the perceptual transition from nonpainful to painful levels of sensation. 6,7However, the question remains whether these responses are attributable to the painful percept or increases in nonspecific effects, such as "orienting or response preparation."For this purpose, we advocate the use of equisalient stimulus designs, where stimuli are carefully matched for nonspecific effects (eg, sensation, unpleasantness, and/or attentional capture).Such designs, while requiring careful instruction and measurement, are necessary for isolating neural responses specifically associated with painful percepts.