Guest Post by David Mehler, Cardiff University and University of Münste
Imagine being able to see and then control your brain activity consciously in real-time. For example, when thinking about a positive life event like your birthday, you would see the activity in areas of your brain’s limbic system increase on a thermometer-like gauge on a screen. Simply seeing this change could further boost your morale: This is the goal of neurofeedback training, which is getting its own boost from a recent research project called BRAINTRAIN.
A consortium of European researchers studying the clinical potential of using functional magnetic imaging (fMRI) for neurofeedback training, BRAINTRAIN in now in its fourth year. The project is working to support clinical trials to test the technique’s efficacy in several focus areas alcohol addiction (Cardiff, UK), anxiety in teenagers (Oxford, UK), autism spectrum disorder (Coimbra, Portugal), obesity (Tuebingen, Germany), and post-traumatic stress disorder (Tel Aviv, Israel).
As a grad student in neuroscience with a medical background, I have been immensely intrigued by a technique with promise for so many differing conditions, especially in the areas of psychiatric and motor rehabilitation. Linking up new technologies with the human brain, I believe, is the future for rehabilitative medicine. As I learned at the annual general meeting of BRAINTRAIN in Coimbra, neurofeedback is one such way to bridge the mind and brain.
Neuroimaging has become a popular tool in neuroscience to investigate mechanisms of various neurological conditions, including those targeted currently by neurofeedback. With neurofeedback, scientists potentially can use this knowledge to help patients increase “healthy” brain signatures or decrease brain signals associated with symptoms like craving in addiction.
In the training process, patients can learn to explore different strategies and the effects they have in self-regulating the brain signal. In doing so, neurofeedback training goes beyond a purely correlational approach. The brain signal, which is normally regarded as the dependent variable in a classical neuroimaging experiment, becomes the dependent variable: The brain activity associated with a certain thought process determines the feedback the person sees. This feedback acts like a teaching signal to learn optimal mental strategies. It is thus an intervention that puts the individual at the core of the experiment.
For example, in a study of alcohol addiction, David Linden and colleagues at Cardiff University are investigating if neurofeedback training effectively helps patients to stay sober. The protocol uses personalized images to provide immersive visual feedback to patients. In one such trial, while in the fMRI scanner, patients are presented on a screen with a picture of a favorite drink; the size of that picture corresponds to the activation level of a brain area involved in the craving sensation. The patient’s goal is to use mental strategies that work well in reducing the craving – for example, remembering situations in which patients remained abstinent. The idea is to pair successful reduction of the feeling with less activation in the respective brain area and reduce the size of the shown picture.
Another exciting application of neurofeedback training uses a more immersive virtual reality (VR) setting. In a study by Gal Raz of Tel Aviv University, the findings suggest that individuals show greater learning effects when being presented with feedback based on VR settings. VR-based feedback is more immersive and thus can boost learning to self-regulate brain activity compared to more simple visual feedback forms such as a thermometer.
For example, one VR setting that has been successfully tested is a busy waiting room of an emergency unit in a hospital. Through relaxation techniques, participants learn to deactivate brain regions involved in anxiety. As they do so, the VR setting changes: the waiting room gets less busy and noisy.
Another avenue I am particularly interested in and hope to contribute to directly through my research is the development of semantic neurofeedback. Rainer Goebel from Maastricht University, presented some exciting recent work including a fMRI spelling device. This device can measure brain activity and then assign unique neural patterns to each letter of the alphabet, thus allowing people to answer questions with their thoughts. This technology could assist in communication for patients with locked-in syndrome – patients who are self-aware but cannot communicate often due to a neurodegenerative condition that affects the brain stem. (A great example of this is new brain computer interface work out of the University of Tübingen.)
With new algorithms, it might be possible to decode mental images in real-time and provide individuals with feedback based on a visual reconstruction of their imagination. For instance, when a patient imagines the number nine, the fMRI spelling device could decode the number and present the decoded image back to the subject. In the beginning, this decoded image might be blurry or ambiguous. But the feedback would allow the individual to optimize the mental imagery of the number nine such that it can be better decoded until the image becomes clear. Our physical eyes would get to see a reconstruction of what our “inner eye” perceives while we imagine an object.
One of the biggest challenges for all these lines of research is that fMRI is a relatively expensive technique that is only available in large clinical and research settings. BRAINTRAIN hopes to transfer the knowledge gained from fMRI studies to more affordable technology, including mobile solutions that involve electroencephalography (EEG), a cheaper, widely available neuroimaging technique.
For example, Talma Hendler’s group from Tel Aviv has recently introduced a new EEG marker that has been developed using machine learning and fMRI data that was simultaneously acquired. This new marker, called “amyg-EFP, is an EEG fingerprint of the amygdala, reliably representing the activity of the emotion center of the brain. This development is very exciting because the amygdala is a deep brain structure and its activity is not detectable with conventional EEG recordings. The amygdala is a promising target for treating post-traumatic stress disorder (PTSD) because it is a central structure in the fear processing network. Currently, the team is conducting a clinical trial in which patients recruited from local hospitals with PTSD symptoms are trained on relaxation techniques while they receive auditory feedback based on the amyg-EFP. Successful relaxation result in a decrease of amygdala activity, which is reflected in a change of the tone patients hear.
Seeing so many potential new application of neurofeedback at the BRAINTRAIN meeting – and in this recent Nature Reviews Neuroscience paper – has inspired me to incorporate many of the findings into my own research, to help patients with neurodegenerative conditions such as Parkinson’s disease. As the outcomes of more BRAINTRAIN trials become available, we will more fully understand the potential of real-time fMRI as a clinical tool for treating patients.
David Mehler is an MD-PhD candidate in medicine and neuroscience at Cardiff University and University of Münster. He uses neuroimaging techniques (fMRI, EEG) to investigate neurofeedback training in healthy participants and patients with a focus on psychiatric and motor rehabilitation. Follow him on Twitter.
Are you a member of CNS with research you want to share? Consider contributing a guest post about your work or trends in the field. It could then become featured on the online magazine Aeon. Email your ideas to CNS Public Information Officer, Lisa M.P. Munoz (cns.publicaffairs[@]gmail.com).