Project Type

URC Presentation



College or School


Class Year



Neuroscience and Behavior

Faculty Research Advisor

Donald A Robin



Chronic pain (CP) is a significant public health problem that impacts over 100 million people in the United States. Treatment of chronic pain costs an estimated $600 billion annually (Gereau et al., 2014). Understanding the underlying brain regions and networks involved in CP can help inform appropriate interventions for this debilitating condition. Neuroimaging meta-analysis is a study technique that allows for the distillation of inter-study coordinate-based data into statistical conclusions of greater power, thus facilitating evidence-based practice. Emerging research in meta-analyses of chronic pain is beginning to recognize the importance of contemporary methodological approaches, such as network analysis. However, few applications of this technique exist in the current literature (Ayoub et al., 2019). The objective of the present study was to identify brain regions active while experiencing pain in patients with CP and in healthy controls (HC) using the activation likelihood estimate (ALE; Eickhoff et al., 2009). This objective functions as an update to previous meta-analytic approaches to chronic pain and healthy controls (Jensen et al., 2016). Subsequent meta-analytic connectivity modeling (MACM; Robinson et al., 2010) was performed with the objective of identifying underlying network co-activation patterns in CP pain activations as a step towards informing therapeutic intervention.


A systematic literature review was conducted for fMRI studies involving CP patients or HC undergoing pain induction via database searching of PubMed and BrainMap (Fox & Lancaster, 2002). Specific inclusion criteria consisted of task-based fMRI in human subjects, whole-brain MRI data, results summarized using stereotactic coordinates in either Talairach or MNI space, and studies published since the year 2000. Talairach coordinates were converted to MNI space via BrainMap software using the Lancaster transformation (Lancaster et al., 2007). Two separate meta-analyses were performed using GingerALE (Eickhoff et al., 2009) to first identify brain regions consistently activated in HC during pain induction and then to identify brain regions consistently activated in CP during pain induction. Peak activation coordinates from both ALEs were used as seed regions for subsequent MACM analyses. Multiple MACMs were performed for each subject group (HC and CP). Importantly, the MACMs were performed to examine region of interest (ROI) co-activations between each ROI and every other ROI in the related networks of either HC or CP subjects.


The literature review yielded 92 studies involving HC undergoing pain induction versus rest conditions and 21 studies involving CP patients undergoing pain induction versus rest conditions. The ALE involving CP (419 subjects, 398 coordinate foci) revealed peak activation in 7 clusters including right medial frontal gyrus, bilateral insula, right postcentral gyrus, left lentiform nucleus, right thalamus, and right claustrum. The ALE involving HC (1453 subjects, 2104 coordinate foci) revealed peak activation in 5 clusters including right claustrum, left thalamus, right cingulate gyrus, right supramarginal gyrus, and left cerebellum. The HC MACMs revealed significant connections among nearly all regions. The CP MACMs revealed significant connections between the right medial frontal gyrus and the insula among other connections.


Decreased pain-related activation of the thalamus, S1, and insula in CP patients compared to pain-related HC data indicates a weakened sensory and interoceptive response to pain in patients, while increased activation in the MFG in CP patients compared to HC data suggests a hyper-aware emotional vigilance to pain in patients. Functional MACMs suggest that CP conditions are associated with aberrant functional network connectivity perhaps in the form of inefficient network processing resulting from thalamic decrease in function and that consistent pain may induce distinct network changes via central sensitization. The neural correlates of pain are complex and are crucial to study in the context of chronic pain to develop novel treatments in the future.


The authors thank, first and foremost, the BrainMap team at the Research Imaging Institute in San Antonio for their extensive and thorough work designing and managing the BrainMap imaging database and associated software suites (Fox & Lancaster, 2002).

NCW_URC.mp4 (4767 kB)
URC Presentation MP4