The eye is negatively charged at the retina in relation to the cornea. It is a dipole. This dipole creates a voltage field. Believe or not, when you move your eyes the electric field recorded at the head changes. This fact has a lot of practical applications, as an example you can follow the point-of-gaze.
There are two kinds of ocular movements: slow and rapid. I would like to propose an experiment. Make another person follow a slow moving pen at 40 cm of his/her eyes. You will see a smooth movement following the object. Now instruct the subject to do the same without object. You will not see the smooth movement of the gaze. You will see the so-called saccadic movements. Even more, the subject will not be aware of this fact. He will feel that he is following the object smoothly. Saccadic movements are the only movements able to be voluntarily controlled. This example is included only to show that different kinds of ocular movements can be controlled by different neural systems.
Fifty years ago it was observed that sleeping persons had intermittent periods of rapid eye movements: REM. Many other physiological functions changed accordingly; so, sleep is not unitary, it includes two different states REM and non-REM. The whole EEG showed important changes in REM sleep. The brain is very active in REM sleep but there is something like a clutch that disconnects brain and body, and the great activity maintained by the brain is not transmitted to the body. Only one voluntary muscular group is free to move: the extraocular muscles. So, during REM sleep, rapid eye movements can be detected by using the same device used to record the EEG. This signal is called electrooculogram (EOG).
One amazing feature. Usually we are not aware that we are dreaming. Some people are able to feel that they are dreaming while they are sleeping. They have no control on their bodies but they retain the control of ocular movements. Even more, they can indicate the researcher that they are dreaming by using ocular movements. This phenomenon is called lucid dreaming.
It is time to return to our subject -hearing physiological recordings- and hear the electrooculogram. Our recording st7121j0.rec contains the signal EOG horizontal. We export it at a 400 times acceleration to the file eog.wav
Following the same procedure than in the EEG case (setting the upper frequency limit at 10000 and the window at 1024) we get something similar to the result shown in figure 5.1. You can compare it with 4.1
In the figure, a REM period has been selected. Notice that the original signal shows an increase in amplitude (the eye movements) associated with REM periods. We can see the appearance of this kind of eye movements in figure 5.2
The rapid eye movements are rapid because they have a high slope. They correspond with saccadic eye movements. But we want to hear their sound. Let's do it with Wavesurfer.
The first surprise is that we can hear the elements of non-REM sleep (spindles and slow waves). It is a very well known feature from long time ago that the EEG can be seen at the EOG signals and as a matter of fact this is the main reason to use two channels for eye movement recording (see Rechschaffen and Kales manual). We can hear the rapid eye movements as scratching noise. Since rapid eye movements appear intermittently during REM sleep, they can be more or less differentiated from wakefulness where this scratching noise sound is more continuous. Even more, each arousal used to be associated with artifact and ocular movements. Arousals are also heard in the EOG (they are seen at the spectrogram as vertical lines covering all the frequencies).
So we could recognize a lot of different features on the EOG. One of the main difficulties in monitoring sleep is that it is hard to place the electrodes on the skull. It is much easier to attach electrodes on the face than on the skull. The EOG electrodes are placed around the eyes and, by hearing this signal, we can get some insight into sleep detection, REM periods, arousals, spindles and slow wave sleep. Not too bad considering that we only used a couple of superficial electrodes attached to the skin.