Sensory Evidence and Expectations in Pain Processing

Overview

Pain is a highly complex and subjective phenomenon which is not only rooted in sensory information but also shaped by cognitive processes such as expectation. However, the interaction of brain activity cording sensory information and expectation in pain processing are not completely understood. Predictive coding models postulate specific hypothesis about the interplay between bottom-up sensory information and top-down expectations in terms of prediction errors and predictions, respectively. They further implicate brain oscillations at different frequencies, which play a crucial role in processing prediction errors and predictions. More specifically, recent evidence in visual and auditory modalities suggests that predictions are reflected by alpha (8-13 Hz) and beta oscillations (14-30 Hz) and prediction errors by gamma oscillations (60-100 Hz). However, for the processing of pain, these frequency-specific relationships have not been addressed so far. The current project aims to investigate brain activity which reflects predictions, prediction errors and sensory evidence in pain processing using a cueing paradigm. To this end, we will apply painful stimuli with low and high intensity to the dorsum of the left hand in 50 healthy subjects. A visual cue, preceding to each painful stimulus, will predict the intensity of the consecutive painful stimulus (low vs. high) with a probability of 75%. After each painful stimulus, participants will be asked to rate the perceived pain intensity. Electroencephalography (EEG) and skin conductance will be recorded continuously during anticipation and stimulation intervals. This paradigm enables us to compare pain-associated brain responses of validly and invalidly cued trials, i.e. the representation of the prediction error, on the one hand. On the other hand, brain activity related to predictions can be investigated in the anticipation interval preceding to the painful stimulus by comparing trials with low and high intensity cues. Further, we will compare models including predictions, prediction error and sensory evidence to ascertain the involvement of each brain response in processing sensory information and expectation. Results of the study promise to elucidate the interplay of predictions, predictions errors and sensory evidence in pain processing and how they differentially relate to neural oscillations at different frequency bands and pain-evoked responses.

Not needed