Self-regulation of brain activity in patients with postherpetic neuralgia: a double-blind randomized study using real-time FMRI neurofeedback

Abstract

Background: A pilot study has shown that real-time fMRI (rtfMRI) neurofeedback could be an alternative approach for chronic pain treatment. Considering the relative small sample of patients recruited and not strictly controlled condition, it is desirable to perform a replication as well as a double-blinded randomized study with a different control condition in chronic pain patients. Here we conducted a rtfMRI neurofeedback study in a subgroup of pain patients - patients with postherpetic neuralgia (PHN) and used a different sham neurofeedback control. We explored the feasibility of self-regulation of the rostral anterior cingulate cortex (rACC) activation in patients with PHN through rtfMRI neurofeedback and regulation of pain perception.

Methods: Sixteen patients (46-71 years) with PHN were randomly allocated to a experimental group (n = 8) or a control group (n = 8). 2 patients in the control group were excluded for large head motion. The experimental group was given true feedback information from their rACC whereas the control group was given sham feedback information from their posterior cingulate cortex (PCC). All subjects were instructed to perform an imagery task to increase and decrease activation within the target region using rtfMRI neurofeedback.

Results: Online analysis showed 6/8 patients in the experimental group were able to increase and decrease the blood oxygen level dependent (BOLD) fMRI signal magnitude during intermittent feedback training. However, this modulation effect was not observed in the control group. Offline analysis showed that the percentage of BOLD signal change of the target region between the last and first training in the experimental group was significantly different from the control group's and was also significantly different than 0. The changes of pain perception reflected by numerical rating scale (NRS) in the experimental group were significantly different from the control group. However, there existed no significant correlations between BOLD signal change and NRS change.

Conclusion: Patients with PHN could learn to voluntarily control over activation in rACC through rtfMRI neurofeedback and alter their pain perception level. The present study may provide new evidence that rtfMRI neurofeedback training may be a supplemental approach for chronic clinical pain management.

Fig 1. Design of the rtfMRI neurofeedback experiment.

A) a scrolling line graph representing BOLD activation shown to patients during scanning. They were instructed to increase the line within the green stage, decrease the line within the blue stage, and keep quiet during gray stage. B) Paradigm of the rtfMRI nurofeedback study. The experimental protocol was consisted of a localizer task, rest, rtfMRI neurofeedback training, pain rating and anatomical scan. During a training run, patients underwent rtfMRI neurofeedback training consisted of alternating blocks of Rest (R, gray block), Increase (I, green block), and Decrease (D, blue block), lasting 30, 60, 60 seconds respectively.

Fig 2. Example of Regions of Interest (ROI) selection for rtfMRI neurofeedback training.

Two squared regions of interest (1 cm per side) were placed for acquiring scrolling line graph. A) rACC was selected based on the activation observed in the localization task. B) PCC was selected according to the anatomic expertise of a senior radiologist. Following radiological notation, the left side (L) of the brain is shown on the right, and the right side (R) of the brain is on the left.

Fig 3. The brain activation maps of rACC by on-line analysis in one patient.

Brain activation of rACC during up regulation in the first training run (A1) and last training run (A2). A stronger activation can be seen in last training run, indicating a successful training.

Fig 4. Results of off-line ROI analysis and NRS change.

Mean BOLD signal change (Δp) and NRS change during up and down regulation between experimental and control group were compared. The symbol * indicates the difference between the two groups is significant at the 0.05 level. Values are mean ± standard error of the mean (SEM).