The potential of fMRI-based biofeedback for chronic back pain ^†

Key points

• Real-time functional magnetic resonance imaging (rtfMRI) neurofeedback is a recently introduced advanced functional MRI technique by which it is possible to learn voluntary control over specific brain areas by means of operant conditioning [1].
• The few currently available rtfMRI neurofeedback studies in healthy volunteers have demonstrated the following fundamental properties of voluntary control of brain activations:

  a.

  The neurofeedback effect is specific and spatially highly localised in the brain [1].

  b.

  A gradual up and down regulation is possible.

  c.

  The neurofeedback effect persists over time [2].

  d.

  rtfMRI neurofeedback may influence behaviour e.g., some aspects of linguistic processing (emotional prosodic) improve after voluntary up-regulation of the right-hemispheric analogue of Broca’s area [3].

• The ability to learn voluntary control over specific brain areas suggests that rtfMRI neurofeedback may be applied to numerous diseases. Currently there have been only two clinical rtfMRI neurofeedback studies showing a positive effect on chronic pain [4] and on tinnitus [5]. rtfMRI neurofeedback is naturally very costly in the initial training phase, but once an individual neurofeedback strategy is learned it can be applied by the individual in everyday situations without additional cost, making this method potentially costeffective in the long term, given that chronic diseases persist over many years.
• With respect to potential clinical application of rtfMRI neurofeedback e.g., in chronic back pain, it is highly probable that greater control over brain activations will result in a more pronounced behavioural effect, which means that the method should be optimised to obtain maximum regulation of brain activations in multiple domains. For example, a motivational feedback consisting of an adapted game may be more efficient than the non-motivational thermometer-style feedback typically implemented at present (activation visually presented as increasing or decreasing bar graph). Other dimensions in need of optimisation are duration and number of feedback sessions, the delay between feedback sessions, “booster” feedback sessions, etc.
• If an optimal and long-lasting effect is to be obtained from rtfMRI neurofeedback, it should be embedded in a multidisciplinary clinical approach.