EOG SIGNAL PROCESSING ALGORITHM USED IN EYE MOVEMENT DETECTION
EOG SIGNAL PROCESSING ALGORITHM USED IN EYE MOVEMENT DETECTION

Summary/Abstract: The EOG signal is based on the potential difference between the cornea and retina (also known as the corneo-retinal potential) and can be measured by using electrodes placed around the eyes. The amplitude of the EOG signal varies between 50 ... 3500 uV. The eye can be seen as a dipole with a positive pole in the retina and the negative one in the cornea. Thus, an electric potential field is created that is changing proportionally with the rotation of the eye. The movements can be assessed from a single eye, but because for a healthy human being the eye movements are coupled, in this paper the signal obtained from both eyes is used. Furthermore, this method allows assessing both vertical and horizontal signal simultaneously. One of the most complex algorithms used to detect steep slopes and peaks within medical signals is the Pan-Tompkins algorithm, but is especially used for ECG signals. This is the reason why the previous studies focused on the detection of the abrupt discontinuities in the EOG signal, from the wavelet transform perspective. In this manner, the points of discontinuity determined by the eye movement will generate high wavelet coefficients, regardless of scale. Starting from previous results regarding different technics used for eye movement investigation and electrooculogram signal (EOG), the paper presents a robust algorithm that is capable of performing real time eye movement detection. This algorithm was implemented in LabVIEW, based on virtual instruments and it was successfully tested in the laboratory of Biomedical Engineering from Technical University of Cluj-Napoca.

• Details
• Contents