Eye Movement Perimetry in Glaucoma Patients

Abstract

 

Purpose

 

We have developed a new perimetry test that evaluates two novel eye movement measures in addition to visual threshold. We hypothesized that optic nerve damage seen in glaucoma patients will result in increased visual threshold, reduced saccadic accuracy, and lengthened saccadic latency when compared with normals.

 

Methods

 

19 glaucoma patients (56 to 78 years) had visual thresholds tested at 3 test locations based on Standard Automated Perimetry (SAP)-determined visual loss: normal sensitivity (+2 to -2dB on total deviation plot), mild visual loss (-3 to -6dB), and moderate loss (-7 to -16dB). 29 age-matched ocular healthy subjects (49 to 74 years) were tested at 3 test locations based on eccentricity: central (3, -3), middle (3, 9), and peripheral (-21, 3). Frequency of seeing curves were constructed using a circular light gray stimulus presented on a dark gray background. Size was varied and location presentation randomized. An EyeLink1000 infrared camera recorded saccadic eye movements to the stimulus. Using threshold and suprathreshold data from frequency of seeing curves, we compared visual threshold (50% seen stimulus), saccadic accuracy (measured as localization error—the distance between recorded final eye position and stimulus position), and saccadic latency (time taken to initiate a saccade to the stimulus) between glaucoma patients and normals matched in eccentricity. Differences among groups were assessed using ANOVA.

 

Results

 

Glaucoma patients had higher visual thresholds (Threshold, graph A, p < .001) and decreased saccadic accuracy (Localization Error, graph B, p = .008) as compared with normals. Eye movement perimetry also showed reduced saccadic accuracy for test locations that SAP considered as normal sensitivity. Threshold increased and localization error increased in this order: normals, glaucoma normal sensitivity, glaucoma mild damage, and glaucoma moderate damage. Saccadic latency measures showed no significant differences between normals and glaucoma patients (Latency, graph C, p = .12).

 

Conclusions

 

Eye movement perimetry shows a significant increase in saccadic localization error with increasing glaucomatous damage; saccadic latencies, however, did not significantly change.

 

 

Diamond designates mean value for normals and glaucoma test locations of normal or mild or moderate damage. Bars designate 1SD. Stars designate statistical significance.

 

Keywords: 525 eye movements: saccades and pursuits • 758 visual fields • 642 perimetry