Background & Aims
Functional magnetic resonance imaging (fMRI) serves as a crucial tool for offering insights into the neurobiological foundations of the pain experience. However, the dynamic nature of pain could present a challenge in precise modeling, since pain is delayed following stimulus onset, often changes during a sustained stimulus, and also takes time to dissipate [1]. This study aimed to improve modeling of stimulus-evoked BOLD signals by integrating the temporal course of pain experiences into fMRI analyses. We hypothesized that by considering the temporal dynamics of pain, we could improve fMRI signal robustness, yielding a more precise representation of the pain in response to stimuli. Specifically, our objectives included 1) accounting for individual subject-by-subject variability in temporal pain delay, 2) accounting for group average temporal pain delay, and 3) applying the average temporal pain delay to a new dataset that included individuals with temporomandibular disorder (TMD).
Methods
Twenty-seven pain-free individuals (14 females, mean age = 23.2 yrs) participated in an initial study involving 20-sec blocks of moderately painful (i.e., 40/100) noxious heat stimuli interleaved with 20 sec OFF blocks. Outside of the scanner, participants continuously reported their perceived pain (captured at 10 Hz) during this paradigm, enabling the calculation of temporal delays in pain responses. Subsequently, participants underwent an fMRI scan in a Siemens 3T Prisma Fit Scanner, where identical noxious heat stimuli were applied in the same ON/OFF block paradigm to capture stimulus-evoked brain responses. We assessed activations by contrasting stimulus ON vs. OFF periods (traditional) and compared this method with contrasting pain ON vs. OFF periods by shifting the ON blocks based on the measured delays in pain onsets. Finally, we applied these delays to the analysis of a different fMRI study in patients with TMD (n = 23) and compared this with the stimulus ON vs. OFF method.
Results
Temporal delays in the onset of pain (i.e. ratings > 5/100) following stimulus onset ranged from 0-17 sec, and the average delay was 8 sec. These findings suggest that modeling brain activations based solely on stimulus onsets and offsets may not be optimal. When considering BOLD signals in key pain-processing regions, including the thalamus, insular, cingulate, and primary somatosensory cortices (S1), incorporating individual delays substantially enhanced the signals compared to the traditional approach. Employing the group average delay in pain instead of each individual’s delay resulted in even greater enhancement of effects. Applying the identified 8-sec delay observed in this initial study to similar data collected in a separate group of individuals with temporomandibular disorder (TMD), the results again demonstrated much more robust BOLD signals in canonical pain-processing areas such as S1, when applying this delay (e.g. peak p value in cluster decreasing from 0.03 to 0.0002).
Conclusions
In conclusion, our study highlights the potential of integrating dynamic aspects of the pain experience into the analysis of noxious stimulus-evoked BOLD signals, resulting in notable improvements in signal detection in experimental settings. Individual-level analysis revealed a diverse range of temporal delays, averaging approximately 8 sec between the onset of a thermal stimulus and the emergence of non-negligible pain. This individualized approach substantially enhanced BOLD signals in canonical pain-processing areas, a effect further amplified when utilizing the group average delay. The improvement in signal strength in patients with TMD suggests that 1) this could also improve fMRI studies in chronic pain populations. In future studies it may be important to measure temporal dynamics of pain in patient groups, since there could be systematic differences across groups like increased painful aftersensations in those with chronic pain [2].
References
1. Hashmi JA, Davis KD. Effect of static and dynamic heat pain stimulus profiles on the temporal dynamics and interdependence of pain qualities, intensity, and affect. J Neurophysiol. 2008 Oct;100(4):1706-15. doi: 10.1152/jn.90500.2008. Epub 2008 Aug 13. PMID: 18701756.
2. Schreiber KL, Loggia ML, Kim J, Cahalan CM, Napadow V, Edwards RR. Painful After-Sensations in Fibromyalgia are Linked to Catastrophizing and Differences in Brain Response in the Medial Temporal Lobe. J Pain. 2017 Jul;18(7):855-867. doi: 10.1016/j.jpain.2017.02.437. Epub 2017 Mar 11. PMID: 28300650; PMCID: PMC6102715.
Presenting Author
Daniel E. Harper
Poster Authors
Daniel Harper
PhD
Emory University
Lead Author
Sofia Heras
Emory University
Lead Author
Zeynab Alshelh
PhD
Emory University; University of Sydney
Lead Author
Alia Lawhorne
BS
Emory University
Lead Author
Maya Iberkleid
Emory University
Lead Author
Ali Alsouhibani
Qassim University College of Applied Medical Sciences, Department of Physical Therapy
Lead Author
Topics
- Pain Imaging