eye movement recording method


The principle of the EOG as an eye movement recording method is demonstrated schematically for the horizontal component in fig.1. The electrodes have to be attached carefully near the eye. The recorded horizontal potential difference UH is a function of the position α of the eye with respect to the electrodes and the CRP, which is assumed to be constant during the recording time. Because the CRP is a slowly changing function of luminance steps the influence of changes in the CRP can be disregarded using nearly constant illumination and repeated calibrations. Metabolic influences e.g. alcohol should be avoided during the recording time. By means of a suitable offset voltage the DC-amplifier can be adjusted so that the output voltage is zero for the the straight eye position (reference compensation).




Although different eye movement recording systems with their advantages and disadvantages exist (electromagnetic search coil method or the eye trackers for example), the EOG is applicable in many cases.


Advantages of the DC-EOG

The EOG is easy to use, also in children and patients confined to bed

The signals are measured with respect to the head

Registration of eye movements with closed eyelids and during sleep are possible

Artifacts from eyelid blinking can be detected easily

Eye movement data are analog and the sampling rate of a following analog-digital-converter (ADC) can be choosen freely

With naso-temporal electrode application the linearity amounts to 20 degrees, with bi-temporal up to 30 degrees

The EOG is the most inexpensive eye movement recording system


Disadvantages of the DC-EOG

Superposition of signals from mimic or chewing muscles (EMG)

Dependency on changes in CRP

DC-recordings always have the problem of drift superimposed on the signal of interest


Overcoming the disadvantages

The influence of EMG artifacts can be reduced by instructing the subject to keep mouth open and avoid innervation of the mimic musculature

Dependency on changes in CRP can be eliminated by reference compensation and repeated calibration (see below)

The drifts can be reduced by careful application of the Ag/AgCl-electrodes and filling them bubblefree with electrode-cream. Technical improvement in EOG amplifiers can reduce the influence of drifts as well. In fig.2 a dual-channel DC-EOG amplifier with automatic reference compensation triggered by hand or computer program is shown. The amplifier is optically isolated to protect the subject from electrical shocks, in accordance with EU standards.




The procedure of reference compensation and calibration is displayed in fig.3. It is initiated by an acoustic signal and the illumination of a central LED. During the reference fixation of this LED the output of the amplifier is set at zero. During compensation the overall gain is coded in the signal. Calibration can be carried out by switching two LEDs at a known position, 20 deg for example.




The linearity of the method is demonstrated in fig.4. Electrodes of each eye are attached nasally and temporally respectively. As can be seen from fig.1 the eye is closer to the nasal than to the temporal electrode with respect to the normal i.e. the straight, position, although eye movements in the nasal direction should show signal saturation earlier than in the temporal. The linear range is also dependent on the anatomical situation. As can be seen from fig.4.




the function is linear up to 20 degrees. Using a bi-temporal electrode attachment the function can be linear up to 30 degrees ( fig.5) and the signal is larger with respect to fig.4.




Some examples of eye movement registrations are attached. Figure 6 is a registration of horizontal and vertical eye movements. In the upper trace the artifacts during blinking are demonstrated. Short involuntary blinks can be detected and voluntary longer lid closures distinguished. During the second voluntary lid closure horizontal eye movements have been recorded.



The second example (fig.7) shows a registration of reading. Fixations and saccades forwards and backwards are shown. This could be used to record reading in legasthenic children and monitor therapy.



In the last example (fig. 8) horizontal eye movements, mostly saccades, are recorded. Using an algorithm separating saccadic from nonsaccadic eye movement components automatically by use of the acceleration signal delivers a membership function which is +1 for saccades to the right, –1 to the left, and zero for all nonsaccadic eye movements. The membership function is plotted in trace 2 and the time after saccades in trace 3. The maxima represent the intersaccadic intervals.




Behrens F., Weiss L.-R. (1992) An Algorithm Separating Saccadic from Nonsaccadic Eye Movements Automatically by Use of the Acceleration Signal. Vision Res. Vol. 32,5, 889-803

---> Back to Homepage

---> Impressum