How to perform neurofeedback
Neurofeedback is a ultimately a training technique with the ability to promote growth and change at the cellular level in the brain. It has been successfully applied in the treatment of depression, anxiety, PTSD, personality disorders, addictions and is effective in disorders of early childhood (such as ADHD) and various education/learning disorders. There is a lot of interest in neurofeedback partially because of it’s potential scope and also because it seems to have the potential to deal with the root causes of issues which are not resolvable by other means. But how do you actually do it?
In the neurofeedback technique, some specific parameters of the EEG signals recorded from a subject’s scalp (such as an EEG power level in a given frequency band, or a ratio of EEG powers in different frequency bands) are presented to the subject (feedback) through visual, auditory, and or a tactile modality with the task to voluntarily alter these parameters in a desired direction. (with the idea of leading to a more efficient mode of brain functioning - it presupposes the brain will learn)
The position of electrodes and the chosen EEG parameters (called neurofeedback parameters) vary depending on the particular goals of neurofeedback. Altogether, the position of electrodes and the neurofeedback parameters define a so-called protocol of neurofeedback.
For medical applications, most practitioners implicitly or explicitly use what Kropotov refers to as a “bulldozer principle” of neurofeedback. According to this principle the aim of neurofeedback is to normalise what is seen as a pathologically abnormal EEG pattern.
So, if there is an excess of some EEG parameter in a particular patient and in particular location in the cortex, the aim of neurofeedback is to train this parameter DOWN, if there is a lack of some other EEG characteristic, the corresponding neurofeedback parameter is trained UP. The method works like a bulldozer filling in the cavities and excavating the bumps in a terrain.
Of course, and this is really important, at the end of this whole process what matters is not the achievement of a normalisation of the EEG but what the clinical or practical outcomes are for the patient/subject.
The current practice of neurofeedback in clinical applications presumes several steps for its implementation. As stated above the first step includes constructing a neurofeedback protocol on the basis of a quantitative electroencephalogram (qEEG) assessment. To do this first step, hardware such as a NeXus 32 system is required - something that supports the many discrete electrode sites.
A tremendous amount of empirical knowledge in EEG analysis reveals some abnormal qEEG patterns associated with various medical and psychiatric disorders.
Placement of at least 19 electrodes is usually required for qEEG analysis. Spectral characteristics of EEG in eyes open, eyes closed conditions, and in some psychological tasks (such as an arithmetic task) are compared statistically to a normative database.
The comparison to the normative data provides scientifically objective (if sometimes controversial) information on how the patient’s spontaneous brain activity differs from age-matching healthy subjects. The statistically significant deviations from normality define the parameters of neurofeedback procedure such as the position of electrodes (e.g., Fz, Cz) and the selected neurofeedback parameter (e.g., beta theta ratio). There are a number of commercially available databases on the market and these should be scrutinised carefully as they may offer advantages and disadvantages depending on the application area.
After defining the necessary electrode positioning we need to define a neurofeedback parameter – a spectral characteristic of EEG that we are going to train.
During neurofeedback training, usually two electrodes are placed on the scalp at locations defined at the first qEEG stage. A ground electrode could be placed on any location, but usually one of the earlobes serves as the background electrode. In comparison to other techniques, such as TMS, DBS, and tDCS no actual electrical current is applied to the brain. It is a measurement process. The electrodes simply measure the ongoing brain EEG whilst the data processing hardware regularly calculates the neurofeedback parameter and presents it in some way to the patient.
The critical points of any programmed neurofeedback procedure include two characteristics that describe:
- how reliable is the computed neurofeedback parameter in indicating the current state of the brain and
- how frequently is the parameter of interest presented to the subject so that the parameter consistently follows changes in the state of the brain.
The first characteristic is actually the integration period within which the EEG current fluctuations are processed. The second characteristic is the time intervals within which the integrated parameter is presented to the subject. From event-related potential (ERP) studies it is known that the state of the brain can be changed in 100 ms time interval whilst the subject reacts to a stimulus within 300 ms. It gives us a temporal estimation of how fast the computed parameter has to be presented to the subject.
With modern hardware such as with the NeXus systems from MindMedia the calculations are rapidly carried out so that theoretically the rate of neurofeedback parameter presentation can be rather high.
When subjects are presented with feedback associated with their brain signals; that is, they see or hear representations of the selected neurofeedback parameter, they are supposed to learn how to change this parameter in a desirable direction.
The subject is usually placed in front of a computer screen. The computer display presented may be very simple - such as a vertical bar on the screen, or can present a kind of complex multi-dimension display with different colours and shapes.
The screen could be also present a computer game designed specifically for neurofeedback, or a commercial computer game in which a parameter of the game is controlled by the neurofeedback.
Another example would be that the subject watches a moviewhilst the neurofeedback parameter calculated by the computer controls the quality of the video picture: the more “normal” the parameter, the better is the picture.
Whatever the method used to present the neurofeedback parameter, the subject concentrates on the feedback and “intends” to learn the association between the parameter and his/ her state. It is not an easy task. Moreover, it can be achieved in different ways. This ambiguity raises a question: how to guide the subject to achieve the task in the most efficient way and how long will it take to learn?
Some practitioners prefer not to give any instructions to their patients by simply saying in effect “Just do it.” Some practitioners give guidance depending on the type of neurofeedback procedure: is it about relaxation or activation? The learning procedure during neurofeedback training may be compared with the technique of how we learn to drive. It takes some time.
Kropotov, Juri D.. Quantitative EEG, Event-Related Potentials and Neurotherapy. 2009
Demos, John N., Getting Started with Neurofeedback. WW Norton & Co. 2005