BOLD Decoding of Individual Pain Anticipation Biases During Uncertainty

A prominent model of pain as a predictive cue posits that anticipation shapes pain transmission and ultimately pain experience. Consistent with this model, the neural mechanisms underlying pain anticipation have the power to modulate pain experience thus understanding pain predictions, particularly during uncertainty, may allow us to ascertain measures indicative of intrinsic anticipation biases. Understanding such biases moves way to precision pain management, as it can guide the individualized treatment. To examine individual pain anticipation biases, we applied machine-learning-based neural decoding to functional magnetic resonance imaging (fMRI) data acquired during a pain-anticipation paradigm to identify individualized neural activation patterns differentiating two certain anticipatory conditions, which we then used to decode that individuals uncertain anticipatory condition. We showed that neural patterns representative of the individualized response during certain anticipatory conditions were differentiable with high accuracy and, across individuals, most commonly involved neural activation patterns within anterior short gyrus of the insula and the nucleus accumbens. Using unsupervised clustering of individualized decodings of anticipatory responses during uncertain conditions, we identified three distinct response profiles representing subjects who, in uncertain situations, consistently anticipated high-pain (i.e., negative bias), subjects who consistently anticipated low-pain (i.e., positive bias), and subjects whose decoded anticipation responses were depended on the intensity of the preceding pain stimulus. The individualized decoded pain anticipation biases during uncertainty were independent of existence or type of diagnosed psychopathology, were stable over one year timespan and were related to underlying insula anatomy. Our results suggest that anticipation behaviors may be intrinsic, stable, and specific to each individual. Understanding individual differences in the neurobiology of pain anticipation has the potential to greatly improve the clinical pain management.