FMRI Deep Phenotyping of Experienced and Imagined Heat-Induced Somatotopy

2510 

Abstract Submission 

Michael Sun¹, Ke Bo¹, Tor Wager²

¹Dartmouth College, HANOVER, NH, ²Dartmouth College, Hanover, NH

Michael Sun  
Dartmouth College
HANOVER, NH

Ke Bo  
Dartmouth College
HANOVER, NH

Tor Wager, PhD  
Dartmouth College
Hanover, NH

Pain perception is a complex neural process that involves encoding somatic nociception, aversive emotions, and contextual predictions about the persistence of stimuli causing pain. While previous research has established that the brain encodes pain differently across body sites and stimulation modalities, the specifics of this encoding – such as location, reliability, and individual differences – remain poorly understood.

This study aims to understand (1) how different body sites vary in warmth sensitivity, heat tolerance, and pain perception; (2) the consistency of subjective pain experiences both within and outside an fMRI environment; (3) differences in brain activity between actual and imagined heat stimulation, as compared to warm stimulation; (4) the brain regions involved in processing experienced versus imagined pain; and (5) the organization of activation within these regions.

The study was pre-registered on the Open Science Foundation platform (https://osf.io/zv4ec). We employed a longitudinal within-subject design of 9 participants (5 male, 4 female). Each participant was assigned individually calibrated High-Heat and Warm-Heat temperatures using an Ascending Method of Limits procedure. Participants then underwent fMRI scanning sessions where a Peltier thermode stimulated a skin-site on their body randomly with either high heat, warm heat, or an instruction to imagine themselves being burned by the thermode. Each participant underwent 40 runs of five repetitions of the thermode stimulating one of eight body sites. At the end of each run, participants were asked to rate the perceived heat intensity, valence (positive vs. negative), and overall bodily comfort.

Experienced and imagined pain deactivated visual and and dorsal attention networks, but only experienced pain recruited the ventral attention network while deactivating the limbic network. Pattern similarity to neurotransmitter maps revealed that imagined pain deactivated regions with high GABA and 5HT receptor density, which is similar to experienced pain. Lexical analysis using Neurosynth topic maps of voxelwise principal components revealed that brain activity was distributed across 3 principal components of Pain Processing (73.94%), Motor Control (17.95%), and Spatial-Sensory Processing (8.12%). Predictably, brain activity across experienced and imagined pain (as well as warm stimulation) were similarly correlated with the first principal component, Pain processing, and similarly anticorrelated with the second principal component, Motor Control (warm stimulation was positively correlated). Experienced and imagined pain diverged on the third principal component, Spatial-Sensory Processing, where imagined pain was positively correlated, and experienced pain was negatively correlated. Warm stimulation was not correlated with this component. Imagined pain was significantly correlated with the Geuter Pain Signature (GPS) and Stimulus Intensity Independent Pain Signature (SIIPS), but not with Neurological Pain Signature (NPS) or Fibromyalgia Pain Signatures (FPS) suggesting that it largely recruits pain-related regions unrelated to nociception.

This study provides new insights into how brain activity is differentiated between experienced and imagined pain from different body sites. These findings have implications for our understanding of the neural basis of pain perception and its variability among individuals. Specifically, there may be a subset of individuals who can activate regions implicated in pain signatures such as the GPS, FPS, NPS, or SIIPS, which may be phenotypes relevant for potentiated pain or chronic pain experiences. The degree to which these signatures map onto reported pain exhibited significant bodysite and individual variability. There are potentially different bodysite-relevant pain patterns per individual, which suggests the need to fine-tune these signatures for individuals.

Imagery

Activation (eg. BOLD task-fMRI) ²

Multivariate Approaches

Univariate Modeling

Perception: Pain and Visceral ¹

FUNCTIONAL MRI

Pain

Perception

Pre-registration

Somatosensory