EEG

I’ve been threatening to do a review of the Pocket Neurobics Pendant since I launched this blog. The time has come for me to not do so. Having used the EEG for a while now, I have come to realise that the EEG instrument is the least interesting part of the whole process.

PN-PendantEEG-LR

The Pocket Neurobics Pendant is a perfectly fine and reasonably priced 2-channel unit, consisting of an amplifier module with standard non-touch connectors for electrode wiring, and a USB receiver for the wireless connection of PC to amplifier.  The build is surprisingly shabby, the amplifier part being housed in an MP3 player carcass (this could have been a feature if some way of using the play/stop buttons could have been devised – remote control of the recording software would have been good).

Let’s start by considering what an EEG measures. Beneath the scalp, past the skull, through the protective dura layers, we find the brain’s cortex. I’m going to make some huge generalisations, and show bias to one particular understanding of how the brain works. Cortical activity is well understood at an electrochemical level, the function of some regions has been established, but the overall function of the cortex holds many mysteries.

The broad concept that appeals most to me is that all sensory input makes it’s way to the cortex, via various filters and modifiers in deeper brain structures, where it combines with other inputs, memories, instincts, mood, etc. to initiate actions commands which are then forwarded to the relevant body parts, again via deeper brain structures. Sometime around cortical processing stage, consciousness is able to access the sense and historical data and make its contribution to the impulse to act. It is this contribution from consciousness that we know as free will.

The brain is composed of billions of specialised cells called neurons. There are a number of different types of neurons, some specific to distinct regions of the brain. For the purpose of this discussion, all we need to know is that there are some that can both receive input from other neurons and stimulate others (most neurons), some that receive stimulation from an external source and stimulate others (sensory neurons), and others that are stimulated by other neurons to activate other cell-types (motor neurons).

A typical neuron has the cell body at one end, with a bunch of ‘docking ports’, called dendrites, that have specialised receptors for specific neurochemicals. Projecting from the other end of the body is the axon, a whip-like extension that finds its way to the dendrites of neighbouring (or distant) neurons to form synapses. Synapses, or more correctly, synaptic clefts, are a minute gap between the end of an axon and a dendrite. The synaptic cleft is a highly controlled chemical mixing vessel. When a neuron ‘fires’, an electrical impulse travels through the axon to the end, where a combination of neurochemicals (serotonin, dopamine, norepinephrine and a bunch of others less commonly discussed) is released into the cleft. These chemicals attach to their corresponding receptors causing the recipient neuron to open or close ion channels in the cellular wall. These ion channels are molecular valves that permit the passage of sodium, potassium and calcium ions to enter the cell body. As ions are charged, this results in the neuron acquiring a voltage potential relative to the outside medium. When a threshold is reached, the neuron ‘fires’. And so it goes on. Several complex mechanisms exist to reabsorb or destroy the residue in the synaptic cleft, leaving it largely ‘pure’ for the next event, however residual chemicals can and do influence mental processes. Selective Serotonin Reuptake Inhibitor antidepressants (SSRIs) are the most common pharmaceutical prescribed to alter synaptic chemical balance.

Not all neurons are in the brain and some axons extend all the way from the skull, via the spinal cord, to internal organs, to every muscle in the body, and from every nerve junction (ganglia). The stomach, for example, has a huge number of neurons, and a special participation for serotonergic system, relating to hunger and satiety, a factor in eating disorders.

What all this leaves us with is a great many neurons firing in extremely complex temporo/spacial relationships. The EEG allows us to see variations in voltage at the scalp which are a function of the many individual neural firings in the few square centimetres below the EEG electrode. Individual neural potentials can be up in the volt range, the massed signal at the scalp is in the microvolt range, being made up of many, many pulses not quite in perfect synchrony, some cancelling, some adding. A clear peak in activity at a particular frequency is indicative that the particular area of the brain is involved in some form of synchronous, or co-ordinated activity. The direct relationship between such activity and emotion/mood/behaviour remains open to research, but there’s a lot of good information providing useful direction for clinical trial, and plenty of leads for enthusiasts to pursue.

The EEG device itself consists of one or more low-noise, high-gain amplifiers, usually with filters to eliminate 50/60Hz and other common environmental noise, along with anything above the usual brainwave range (below 100Hz, some EEGs filter at 75Hz, some lower). The input to the amplifier is either via a “wet” electrode system, commonly silver with silver chloride paste, or active electrode, where an amplifier is located right at the scalp, improving noise rejection and contact performance. Good connections are the key to good EEG, and an assistant makes all the difference for self-experimentation.

The output of the amplifiers traditionally went to a mechanical recorder, the signal driving a swing pen-arm, leaving a trace on a moving strip of paper. These days the signals are digitised (resolution in bits ranges from 8 to 16 in ‘amateur’ devices, more in clinical devices) and processed via software.

A trained clinician can interpret a great deal from a raw trace. A layperson such as myself benefits from the incredible analysis and display capabilities of modern software. I have glanced at some of the other EEG software, but I am totally satisfied with BioExplorer. BE uses an object-based interface to create ‘designs’ that interconnect input, processing and output devices on screen. A good range of basic designs are included, so a newby can be up and running quite quickly.

The most immediately useful designs use fourier analysis to sample a slice of time (epoch)  from the raw signal, consisting of a spectrum of frequencies in various proportions, and calculate the power distribution of each frequency within that time period. This is commonly displayed in  ‘spectrum analyser’ format, with each frequency range ‘binned’ into a graph column. Another view is the 3-D or waterfall spectrum – this is my favourite as it makes the various types of activity superbly conspicuous to the untrained eye (see screenshot below).

1020System

Crucially important to the use of EEG is knowing where to take measurements. The international standard for electrode placement is the 10-20 system which divides the cranium into regions determined by percentages of the distance between common ‘landmarks’. Accurate placement is essential only for sharing data and comparing results over time or between individuals. For general experimentation and simple biofeedback, near enough is good enough. Most enthusiast recordings will be based on relative measurement – more or less activity at a particular frequency. A specialised form of EEG, quantitative EEG (QEEG), takes absolute measurements at very precise locations and can be used as an objective diagnostic tool and to accurately quantify response to therapy. QEEG uses calibrated multichannel (64, 128, 256 channels)  instruments, beyond the reach of most enthusiasts. Very little diagnostic information can, or should be, deduced from one or two channel recordings, however a great deal of insight can be gained by enthusiasts through such measurements. Basic neurofeedback can be achieved with a single channel.

The 10-20 system names locations according the brain structure underlying them, plus a number indicating left or right hemisphere and distance from the centreline of the skull. For example, Pz is over the centre of the parietal lobe, P3 is a little to the left, and P4 is a little to the right. Odd numbers are left, even right, and higher numbers are closer to the centreline. A simple 10-20 layout is shown – imagine a map for 256 channels! Ground or reference connections are commonly made to one or both earlobes. Earlobe and scalp electrodes are shown below.

PN-Electrodes-LR

To know where to connect you need to know what you’re hoping to measure, and which brain structure is typically involved. The obvious ones are the temporal region, where evoked/frequency following responses to sound will most likely be found and the occipital region, where responses to light are easily found. Over the central region you’re likely to find sensorimotor (SMR) activity and the frontal lobe is the place to look for mood and thought related activity.

Bioexplorer-00-LR

 The screenshot above shows BioExplorer with a single channel design, routed through numerous Dominant Frequency filters to show the dominant frequency within each range. Raw EEG is shown by the blue trace, and the 3D graph shows frequency power distribution against time. This session was recorded with a single electrode attached to Cz – top of the head over the central region (note that there is no central lobe, C is more or less between parietal and frontal lobes) and gives a good general indication of simple entrainment response. I was running a standard alpha MWS sound/light session, as can be seen by the string of peaks along the 10Hz marker. The activity at the low end is mostly movement noise.

I find performing EEG recordings sufficiently frustrating without a practiced assistant to avoid doing them as far as possible. I tend to do ‘thought experiments’, visualising what I know of the brain and its response to stimulus to experiment initially with just sound and light, cultivating a particular ‘state of mind’ or ‘mode of thought’ before hooking up to the EEG to see if things are happening according to my expectations, and modifying practice accordingly. Largely due to lack of an assistant I have done very little with neurofeedback, preferring to do occasional spot checks to see if I’m on track, rather than using the EEG to control stimulus. There’s a few game-like designs included with BioExplorer that allow you to concentrate on different types of mental activity to cause various onscreen events to take place – similar to biofeedback software such as Mental Games (ThoughtStream) or The Passage (Wild Divine).

Mind Workstation Enterprise can interface with BioExplorer (and some other EEG packages) via Biofeedback Engines, with example sessions for alpha relaxation and to stimulate SMR and Beta activity. MWS EEG engines provide easy access to most BioExplorer parameters, which can be used to control virtually any MWS parameter (see my post “MWS Engines”).

EEG is a fabulous educational tool, and essential for the best implementation of some therapeutic AVS protocols (notably ADD/ADHD and depression). Even with the relatively low-cost ‘home’ EEGs now readily available, it’s expensive to get into. Decent software, such as BioExplorer, can cost as much as the EEG device itself. There is an OpenEEGproject with circuit designs, some kit support and open licensing, but anything other than a commercial package is going to hugely increase the already significant learning curve.

The expectation that signs of large scale entrainment will be easily observed is likely to leave many newcomers quite disappointed. The brain is far too sophisticated to give us simple cause/effect results at the scalp, and it is only by relating scalp measurements to data acquired using more advanced techniques, such as fMRI, and invasive techniques, such as microelectrode single neuron EEG, that real understanding can evolve. Nevertheless, I find it truly amazing that such powerful tools, formerly found only in the clinical setting, are now accessible to enthusiasts like us.

I hope this very lightweight overview does something to fill the gap between lay-knowledge and the bulk of rather heavy information available in books and on the net.

Cheers,
Craig

Advertisements
Post a comment or leave a trackback: Trackback URL.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: