Using functional magnetic resonance imaging (fMRI), researchers have measured an individual’s brain activity to assess pain arising from sensory input with clinically useful accuracy. The findings of the study, published in The New England Journal of Medicine, may provide a method for developing neuroimaging-based diagnostic measures for clinical pain, helping physicians to identify and treat pain and distress.
“Pain in others is hard to relate to because it is intangible and it is often ignored, minimized, or explained away as ‘not real,’ and some groups, including women and minorities, are systematically under-treated,” Tor D. Wager, PhD, lead author and director of the Cognitive and Affective Control Laboratory at the University of Colorado, Boulder, said in an interview. “We need objective measures to corroborate the presence and intensity of pain.”
The study included 114 participants and four separate studies. In the first study, Wager and colleagues used a machine-learning algorithm to identify a pattern of fMRI-based brain activity that was associated with heat pain.
The researchers found that the pattern, which they called nociceptive pain signature, tracked the intensity of a participant without prior knowledge about what each individual’s “pain map” looked like.
“This allowed us to estimate the magnitude of activity in the pain signature pattern and accurately predict pain experience,” Wager said. “The signature accurately tracked increases in pain elicited by slightly hotter stimuli — for example it was 90 percent to 100 percent accurate when stimuli differed by 1°C — when the stimuli were painful, but not when they were warm but not painful.”
Next, they tested the sensitivity and specificity of the signature when compared with warmth and found that the signature was specific to physical pain. The pain signature did not respond to other salient events, Wager said, including anticipatory anxiety related to pain or thinking about past pain.
In the third study, the signature was tested compared with social pain, which activates similar areas of the brain.
“It did not respond to stimuli that elicited social pain and other emotions — in particular, images of ex-relationship partners in individuals who were recently rejected in love,” Wager told Diagnostic Imaging. “The signature was 90 percent to 100 percent accurate at discriminating pain from these other types of events, depending on the test.”
Finally, the researchers tested the signature for responsiveness to the analgesic agent remifentanil, and found that it did respond to the opiate, which is similar to morphine.
The findings were surprising, Wager said, because the researchers did not initially think that fMRI data were reliable enough to track pain with such a high degree of accuracy.
“We initially expected the signature for pain to be different for each individual, which would make it difficult to develop a brain-based pain signature without extensive work in each individual tested and a sample of data where the ground truth about a person’s pain is known,” Wager said. “The fact that it is both sensitive and specific to pain, and is consistent across individuals, suggests that it might be possible to identify brain patterns reliably that are related to multiple types of pain and other emotional experiences.”
Wager believes that among other things, this pain signature could be used to confirm pain in people who cannot accurately report pain themselves, such as the very old, very young or the cognitively impaired. However, he strongly cautioned that it not be used as a way to disconfirm the presence of pain, or as a way to deny treatment because some individuals may have pain that is not captured by the pattern identified in this study.
“There are many conditions in which pain does not seem to directly stem from nociceptive —potentially tissue-damaging — stimulation,” Wager said, “and there are reasons to believe the current signature could fail to detect pain in these conditions, despite the fact that patients are truly in pain.”