How are brainwaves recorded and viewed on a computer?

One way to do this is to record in-vivo (this is usually done with animals although in some unique cases such as when deep brain stimulation [1]).  This means that you cut a small hole in the skull above the brain area you are interested in and then lower recording electrodes into the brain.

In order to put in the electrodes you use a devise called a stereotaxic which holds the anesthetist animal motionless and allows for precise measurements.

This is a sterotaxic used for rats.

Once in the stereotaxic measurements can be taken from two places depending on which is closer to the intended brain area.  These two areas are called Bregma and Lambda which are intersection points of sutures on the top of the skull.  In literature it is reported that, for example, recording was done -1.5 mm anterior (towards the front of the skull), 0.5 mm lateral (away from the middle of the skull) bregma.


Once electrodes are lowered into the brain the signal is passed through a series of amplifiers and filters (to filter out noise such as the 60 Hz band most electronics give off) to a computer.  There are several programs that can be used to capture and analyze the brain signals.  I use Spike 2 (http://www.ced.co.uk/pru.shtml), another common program is Plexon (http://www.plexon.com/plexon_products.html#Software).



Above is an example of what a brain recording looks like.  This example is of simultaneous recordings from the right and left barrel cortex.  The bottom (x-axis) is time and the left axis has four channels the top two from one electrode in the left barrel cortex and the bottom two from an electrode in the right barrel cortex.  Each signal from an electrode is filtered two different ways in order to see neurons spiking (filters at 250-5000 Hz) and local field potential (LFP) (filtered at 0.1 to 100 Hz)  [2].  The local field potential reflects the voltage in the local extracellular space.

The two periods of blocks you can see in the image are showing the response to stimulation of the rat whisker pad (which projects to the barrel cortex).

On closer inspection (image below) you can see the stimulus and the resulting action potential indicating that a neuron is firing in response to the stimulation.  You can see neuron responding to the stimulus in channels 4 and 8 and the LFP response in channels 5 and 9.


One further step you can take to look at a spiking neuron is to take a Stimulus Trigger Waveform Average (STWA).  This means that you average the neurons response to the stimulus over the entire period it is being stimulated.  In  this case there are 100 stimuli separated by 330 ms (3Hz stimuli for 33.3 s).  This results in:


This is, in a sense, a picture of a neuronal response (a putative pyramidal (exitatory) cell based on the duration of the action potential.)

[1] http://en.wikipedia.org/wiki/Deep_brain_stimulation
[2] http://en.wikipedia.org/wiki/Local_field_potential