Neurofeedback with fMRI: A critical systematic review
Introduction
In recent years, neurofeedback using fMRI (fMRI-nf) has increasingly captured the interest of scientists, clinical researchers, practitioners, and the general public. This technique provides individuals with near real-time feedback from their ongoing brain activity (Fig. 1). FMRI-nf offers many advantages over traditional, albeit increasingly challenged, forms of neurofeedback aiming to entrain and control electroencephalographic signals (EEG-nf; Birbaumer et al., 2013). Unlike EEG-nf, fMRI-nf provides millimetric spatial resolution and consistently guides participants to successfully regulate their brain activity indexed by the blood-oxygen-level dependent (BOLD) signal (Thibault et al., 2015). In addition, research on fMRI-nf improves on many key methodological shortcomings that plague typical EEG-nf experiments (e.g., Arnold et al., 2013, Thibault and Raz, 2016)—employing more rigorous control conditions (e.g., sham neurofeedback from an unrelated brain signal) and measuring both learned regulation of the BOLD signal as well as behavioral response. Here we offer a critical systematic review of the fast growing literature on fMRI-nf, with an eye to examining the underlying mechanisms, observable outcomes, and potential therapeutic benefits.
The present review gathers findings from nearly all available primary experiments involving fMRI-nf, which aim to train neural regulation or modify behavior (we exclude case studies and other experiments that present only individual level analyses). We opt for a systematic review rather than a meta-analysis due to the wide variety of experimental designs and statistical methods used in fMRI-nf. Whereas meta-analyses generally focus on a specific treatment and outcome measure, the spectrum of fMRI-nf studies hardly renders itself to this meta-analytic approach—the studies train distinct brain regions, employ a variety of controls, use different time points as their baseline, measure diverse behaviors, and vary in the length of training and instructions provided. While we encourage meta-analyses for more specific questions concerning fMRI-nf (e.g., Emmert et al., 2016), a comprehensive meta-analysis would risk misrepresenting the heterogeneity of the field by assigning a single valuation to the technique as a whole (Moher et al., 2009, Thompson, 1994).
After outlining the parameters of our literature search, we present the distribution of control conditions and experimental designs throughout the field. We then examine the effectiveness of fMRI-nf protocols in (1) training self-regulation of the BOLD signal and (2) modifying behavior. Some scholars speciously conflate these two distinct outcome categories, assuming that altered BOLD patterns will inevitably or necessarily drive observable changes in behavior; however, this assumption hardly holds true. After considering the observable outcomes, we evaluate the status of fMRI-nf as it begins to edge towards clinical acceptance. We conclude that fMRI-nf presents a reliable tool for modulating brain activity, but that current experimental protocols vary too widely to reify therapeutic efficacy and endorse practical guidelines at this time.
Section snippets
Review protocol
We searched the Topic: (neurofeedback) AND (fMRI OR “functional magnetic resonance imag*” OR “functional MRI”) across All Databases and all years in Web of Science on August 25th, 2017 (see Fig. 2 for a flow chart of study inclusion). Of the 434 published articles written in English that were returned, we omitted 114 not directly related to fMRI-nf (e.g., performed neurofeedback with a different imaging modality or used fMRI as a means of analysis only), 72 conference proceedings or abstracts,
Experimental design in fMRI-nf
How does the fMRI-nf literature stack up to the gold standard of experimental science across most clinical research domains: placebo-controlled and double-blind? Ideally, control groups receive a highly comparable treatment that omits the active ingredient or mechanism of action purported to drive improvement, and neither participants nor experimenters can identify who receives veritable versus placebo treatment. Increasingly, fMRI-nf experiments are rising to this standard and employing a
BOLD self-regulation
The question at the heart of fMRI-nf research is whether individuals can learn to volitionally modulate neural activity in circumscribed brain regions. The cumulative evidence suggests that participants can indeed successfully modulate the BOLD signal from a wide variety of brain regions (Fig. 5A). While this overarching finding may spark enthusiasm, we would do well to remember that participants in thousands of imaging studies before the advent of neurofeedback had already regulated their own
Behavioral self-regulation
The promise of fMRI-nf stems from the potential to regulate brain processes and, in turn, to improve well-being. Nonetheless, we remain far from establishing causal links between circumscribed patterns of brain activity and complex human behaviors. Whereas neuroscientists have successfully mapped discrete stimuli onto the sensory cortices (e.g., primary motor, sensory, or visual areas), the neural correlates of psychiatric conditions and multifaceted mental processes appear to rely on the
Sustainability, transferability, and practicality of fMRI-nf
While positive findings abound in fMRI-nf research, the clinical feasibility and value of this technique remains unconfirmed. A few years ago, several prominent neurofeedback researchers stated in an authoritative review that the “real usefulness [of fMRI-nf] in clinical routine is far from being demonstrated” (Sulzer et al., 2013a). The present review suggests that their statement remains valid: to date, few studies have tested clinical significance, examined patient populations, or
Steps forward in neurofeedback protocols
Since the inception of fMRI-nf in 2003, research on neurofeedback has progressed significantly. For one, fMRI-nf makes several important advances over more traditional, EEG-based, approaches to neurofeedback. EEG-nf experiments generally involve dozens of training sessions and often neglect to directly measure whether participants learn to modulate neural activity. In contrast, fMRI-nf requires only a few runs to impart BOLD modulation, and relevant experiments almost always measure neural
Conclusion
The present comprehensive review suggests that fMRI-nf may develop into a powerful biobehavioral intervention. Experiments repeatedly demonstrate that real-time feedback allows individuals to modulate the BOLD signal from a plethora of cortical regions. And yet, BOLD self-regulation falls short of implying behavioral self-regulation. Our in-depth review reveals three important lacunae in the domain of fMRI-nf:
First, replications remain sparse. Of the 99 experiments we identified, few show
Funding
This work was supported by the Canada Research Chair program and Canadian Institutes of Health Research (MOP-106454), Discovery grants from the Natural Sciences and Engineering Research Council of Canada (NSERC) (38656-210) and the Bial Foundation (118/14). Robert T. Thibault, also a Bial recipient, acknowledges an Alexander Graham Bell Canada Graduate Scholarship and Michael Smith Foreign Study Supplement from NSERC. Michael Lifshitz acknowledges a Canada Vanier Scholarship from NSERC and a
Contributors
RTT, the guarantor of this article, led the systematic review, defined the search criteria, extracted data from the publications, and prepared the initial draft of the manuscript. AM and RRR helped review the publications and edit the manuscript. AR and ML provided comments throughout manuscript preparation and worked with RTT to produce the final draft.
References (200)
- et al.
Learned regulation of brain metabolism
Trends Cognitive Sci.
(2013) - et al.(2006)
- et al.
The respiration response function: the temporal dynamics of fMRI signal fluctuations related to changes in respiration
NeuroImage
(2008) - et al.
Hemodynamic scaling of fMRI-BOLD signal: validation of low-frequency spectral amplitude as a scalability factor
Magn. Reson. Imaging
(2007) - et al.
A purely confirmatory replication study of structural brain-behavior correlations
Cortex
(2015) - et al.
Volitional control of anterior insula activity modulates the response to aversive stimuli. A real-time functional magnetic resonance imaging study
Biol. Psychiatry
(2010) - et al.
Regulation of anterior insular cortex activity using real-time fMRI
NeuroImage
(2007) - et al.
NeuroImage Investigation of fMRI neurofeedback of differential primary motor cortex activity using kinesthetic motor imagery
NeuroImage
(2012) - et al.
Decoded fMRI neurofeedback can induce bidirectional confidence changes within single participants
NeuroImage
(2017) - et al.
Learned regulation of spatially localized brain activation using real-time fMRI
NeuroImage
(2004)
Publication bias in clinical research
Lancet
Continuous vs. intermittent neurofeedback to regulate auditory cortex activity of tinnitus patients using real-time fMRI - a pilot study
NeuroImage Clin.
Meta-analysis of real-time fMRI neurofeedback studies using individual participant data: how is brain regulation mediated?
NeuroImage
Real-time fMRI links subjective experience with brain activity during focused attention
NeuroImage
Control of nucleus accumbens activity with neurofeedback
NeuroImage
fMRI neurofeedback of higher visual areas and perceptual biases
Neuropsychologia
Dynamic reconfiguration of human brain functional networks through neurofeedback
NeuroImage
Effects of salience-network-node neurofeedback training on affective biases in major depressive disorder
Psychiatry Res. Neuroimaging
Reduction of cue-induced craving through realtime neurofeedback in nicotine users: the role of region of interest selection and multiple visits
Psychiatry Res. - Neuroimaging
Modulation of functional network with real-time fMRI feedback training of right premotor cortex activity
Neuropsychologia
Connectivity-based neurofeedback: dynamic causal modeling for real-time fMRI
NeuroImage
Self-regulation of the anterior insula: reinforcement learning using real-time fMRI neurofeedback
NeuroImage
Real-time fMRI-based neurofeedback reinforces causality of attention networks
Neurosci. Res.
Brain-machine interface via real-time fMRI: preliminary study on thought-controlled robotic arm
Neurosci. Lett.
What does the HAMD mean?
J. Affect. Disord.
Volitional reduction of anterior cingulate cortex activity produces decreased cue craving in smoking cessation: a preliminary real-time fMRI study
Addict. Biol.
Brief breath holding may confound functional magnetic resonance imaging studies
Hum. Brain Mapp.
The future of psychiatric research: genomes and neural circuits
Science
Real-time fMRI neurofeedback in adolescents with attention deficit hyperactivity disorder
Hum. Brain Mapp.
Empirical neuroenchantment: from reading minds to thinking critically
Front. Hum. Neurosci.
Absence of evidence is not evidence of absence
BMJ Br. Med. J.
Learning to associate orientation with color in early visual areas by associative decoded fMRI report
Curr. Biol.
EEG neurofeedback for ADHD: double-blind sham-controlled randomized pilot feasibility trial
J. Atten. Disord.
Training efficiency and transfer success in an extended real-time functional MRI neurofeedback training of the somatomotor cortex of healthy subjects
Front. Hum. Neurosci.
Visual motion imagery neurofeedback based on the hMT+/V5 complex: evidence for a feedback-specific neural circuit involving neocortical and cerebellar regions
J. Neural. Eng.
Self-modulation of primary motor cortex activity with motor and motor imagery tasks using real-time fMRI-based neurofeedback
NeuroImage
Modulation of functionally localized right insular cortex activity using real-time fMRI-based neurofeedback
Front. Hum. Neurosci.
Improvement in precision grip force control with self-modulation of primary motor cortex during motor imagery
Front. Behav. Neurosci.
Direct instrumental conditioning of neural activity using functional magnetic resonance imaging-derived reward feedback
J. Neurosci. Off. J. Soc. Neurosci.
Real-time neurofeedback using functional MRI could improve down-regulation of amygdala activity during emotional stimulation: A proof-of-concept study
Brain Topogr.
Statistical rigor and the perils of chance
eNeuro
Power failure: why small sample size undermines the reliability of neuroscience
Nat. Rev. Neurosci.
Volitional regulation of the supplementary motor area with fMRI-BCI neurofeedback in Parkinson's disease: a pilot study
Self-regulation of anterior insula with real-time fMRI and its behavioral effects in obsessive-compulsive disorder: a feasibility study
PLoS One
Sustained reduction of nicotine craving with real-time neurofeedback: exploring the role of severity of dependence
Nicotine Tob. Res.
Outpatient Charge Data CY 2014
The weirdest brains in the world
Behav. Brain Sci.
fMRI-based robotic embodiment: controlling a humanoid robot by thought using real-time fMRI
Presence Teleoperators Virtual Environ.
Cognitive and neural strategies during control of the anterior cingulate cortex by fMRI neurofeedback in patients with schizophrenia
Front. Behav. Neurosci.
Multivoxel neurofeedback selectively modulates confidence without changing perceptual performance
Nat. Commun.
Cited by (153)
A brief real-time fNIRS-informed neurofeedback training of the prefrontal cortex changes brain activity and connectivity during subsequent working memory challenge
2024, Progress in Neuro-Psychopharmacology and Biological PsychiatryEffects of neurofeedback training on performance in laboratory tasks: A systematic review
2023, International Journal of Psychophysiology