Development of a test grid using Eye Movement Perimetry for screening glaucomatous visual field defects

Background: Eye Movement Perimetry (EMP) uses Saccadic Eye Movement (SEM) responses for visual field evaluation. Previous studies have demonstrated significant delay in initiation of SEMs among glaucoma patients in comparison with healthy subjects. The aim of the current study was to develop an EMP-based screening grid to identify glaucomatous visual field defects. Methods: An interactive test consisting of 36 locations and two stimulus contrasts (162 cd/m2 and 190 cd/m2 on a background of 140 cd/m2) was evaluated in 54 healthy subjects and 50 primary glaucoma patients. Each subject was presented a central fixation target combined with the random projection of Goldmann size III peripheral targets. Instructions were given to look at each peripheral target on detection and then re-fixate at the central fixation target while the saccades were assessed using an eye tracker. From each seen peripheral target, the Saccadic Reaction Time (SRT) was calculated for contrast level 162 cd/ m2. These values were used to plot Receiver Operating Characteristic (ROC) curves for each test locations and the Area Under the Curve (AUC) values were used to identify the locations with highest susceptibility to glaucomatous damage. Each stimulus location with an AUC less than 0.75 along with its mirrored test location around the horizontal axis were eliminated from the grid. Results: The mean age was 48.1 ± 16.6 years and 50.0 ± 14.5 years for healthy subjects and glaucoma patients respectively. A significant increase of SRT values by 76.5% (p < 0.001) was found in glaucoma patients in comparison with the healthy subjects. From the ROC analysis, ten out of 36 locations meeting the cut-off criteria of AUC were eliminated resulting in a new grid containing 26 test locations. SRT values were significantly different (p < 0.05) between the healthy subjects and glaucoma irrespective of the grids used. Conclusions: The present study resulted in a screening grid consisting of 26 locations predomin