Physio Methods I:
Neurofeedback (EEG & HEG)
What is bio-/neuro-feedback?
Neurofeedback and biofeedback are techniques based on the principle of Operant Conditioning. The development of this psychological concept started with Thorndike (1911/1999) who put forth the idea of the “Law of Effect,” Which says response that produce satisfying effects for the subject become more and more likely to recur. The same holds true for the converse; responses that produce dissatisfying effects for the subject become less and less likely to recur.
B.F. Skinner in the 1940s brought more to bear on the concept of operant conditioning. His formulation goes as follows:
Positive or pleasing feedback increases preceding behavior (and in our case, brain-level or heart-level responses)
Negative or unpleasing feedback decreases preceding behavior
Think of feedback, whether points in a game, or tones, or a vibrating stuffed animal, as a signal that the brain uses to “listen” to itself functioning and shift accordingly. The more reinforcement feedback the brain receives for a given selected metric (the most basic example being power in a specific frequency band), the more the brain will function along those lines. The more the brain will adapt to the different way of functioning.
What is EEG?
Electroencephalography, or EEG, is an old but remarkably useful technology. Berger was the first to discover it in Austria when he first witnessed coherent electrical waves called alpha waves emanating from the back of the head. The technology — which has been spectacularly improved over the years since Berger’s time — works as follows:
Electrical leads, or electrodes, are attached to the client’s scalp. These pick up electrical activity emanating from part localized underneath the electrode sensor. These signals are then sent over through an amplifier, which “magnifies” the signal to a level discernible by a computer, over to the computer software system. That software then massages the data in real time, performing various transformations on the incoming signal. The program then, with the specific protocol’s parameters, adjusts an output feedback signal which is transduced into either visual (video game or visualization), auditory (tones, or volume associated with a visualization) or tactile sensations (vibrating teddy bear). The client receives these feedback sensations and the brain — this is still the mysterious part — figures out a way to integrate all of that information into modifying itself.
The most basic transformation of EEG signals, which serves as a foundation for many if not most modern day neurofeedback protocols, is the Fourier Transform. This transformation takes the signal coming from the amp and converts the raw data into time-frequency data. What does that mean? It basically means that you get a picture of what frequencies are doing as a function of time, usually for the sake of neurofeedback this goes from ~.5Hz to about 45Hz. How this plays into feedback is demonstrable in the simplest and likely most widely utilized protocol class: single band reward. The spectrum derived in the Fourier Transformation is broken down into what some call the “Canonical Bands” delta, theta, alpha, beta, and gamma, and often further down into “upper” and “lower” or indexed based on frequency. For example, alpha is often broken up into “upper” (10-13Hz) and “lower” (8-10Hz) because these two frequency regimes have differentiable psychological and physiological mechanisms. In a single band reward, the client’s real time brain data is Fourier transformed, a specific band of frequencies is evaluated for the power (~V^2 dependence) within that band, and the feedback signal reflects whether the brain is operating above a specified threshold or not. By training up the EEG power in that band, the brain begins to “notice itself” and shift its functioning.
Placements for the electrodes are very important. In EEG-based neurofeedback, where the electrodes are placed on the scalp can mean the difference between an effective session or a session that produces (very temporary) unfavorable symptoms. Different regions of the cortex correspond with different behavioral and psychological functions. For example, a great deal of foresight, planning and sequential thinking is carried out by the left frontal area. By working with specific placements of electrodes, we can titrate responses and capacities based on your goals, challenges and preferences.
We could go into depth about the different regions, from lobes to smaller areas, that correspond to different sets of behaviors and capacities, but suffice it to say there are 12 broad regions, 6 on each side, to take note of: prefrontal, frontal, sensorimotor, temporal, parietal and occipital. These correspond — very broad brush — to planning, execution, motor and sensing, audition and memory, body awareness and visual processing, respectively.
This of course is a vast oversimplification and says nothing particularly useful about the flow of information between regions, “collaboration” between regions, or anything like that. But this oversimplified model conceptually will suffice for an introduction.
What is NIRS based Neurofeedback or HEG (hemoencephalography?
There is a phenomenon called cerebrovascular coupling (CVC), the delivery of blood flow to regions that need activation for the brain to perform some task. The CVC has been shown in studies and clinical experience to be trainable as a feedback modality. Specifically, the prefrontal and frontal areas most associated with the “more human” functions can be trained so that those regions can “burn brighter” by increasing the metabolic turnover in a relatively localized area. By training the brain’s ability to direct blood flow to the region, you improve the metabolic capacity of that area and thus the underlying cognition associated with it. How do we measure the blood flow to make a feedback metric?
We don’t have to go do a fMRI, we don’t have to spend that kind of money. The solution is simple: Near Infrared Spectroscopy or NIRS. It is also referred to as hemoencephalography or HEG neurofeedback. This process involves an infrared LED module and an adjacent sensor that picks up reflected IR light. Oxygenated hemoglobin in the blood reflects the near infrared light impinged on it. NIR luckily passes through the tissue and bone of the skull, making it possible for the sensor adjacent to the LED to pick up the blood flow dynamics under the sensor. In our case, we have this handy NIR sensor-LED strap that fits comfortably across the front of the head, where many studies have shown positive effects of this flavor of neurofeedback. The data signal coming from the sensor module, through the amplifier to the computer system then gets processed by the software and shuttled to the computer game, LIFE, which operates based on the increase, decrease and persistence of blood flow changes. Feedback is “encoded” for the brain in the number of points you gain and a metric called the “attention index”.
What is a mini-Q? The mini-Q is the process and methodology that a lot of practitioners use to get a sense of what is going on within the client’s brain. “Q” is a stand in abbreviation for QEEG which is suggestive of the quality of the process. Q refers to the fact that the EEG recording is compared quantitatively with a full several-thousand-subject-strong database of EEG metrics recorded over the years. This comparison with a database of metrics is not diagnostic, but rather the insights that are gleaned from the comparison are descriptive or suggestive. For example, there are some EEG signatures that have been shown to correspond with a high degree of correlation (think ~85-90%) with specific challenges or complaints. The Q’s comparison with the database is not bulletproof, but it definitely a good look into the client’s neurological landscape.