An objective evaluation of gaze tracking in Humphrey perimetry and the relation with the reproducibility of visual fields: a pilot study in glaucoma.

Abstract

To develop a novel method to evaluate gaze tracking (GT) results and to examine their relationship with test-retest reproducibility of visual field (VF) measurements. Subjects comprised of 42 eyes of 42 glaucoma patients. Vision fixation during VF tests with the Humphrey Field Analyzer was evaluated using the gaze fixation line chart at the bottom of the VF printout. We defined some GT parameters as follows: average tracking failure frequency per stimulus (TFF), average frequency of eye movements between 1° and 2°, 3° and 5°, and more than 6°. Humphrey VFs (24-2 and 10-2 Swedish Interactive Threshold Algorithm [SITA] standard) were prospectively examined twice within a period of 3 months in 42 glaucoma patients. Mean absolute variability of total deviation (TD) values in the test-retest VFs was measured and its relationship to fixation losses (FLs), false positives (FPs), false negatives (FNs), mean deviation (MD), and pattern standard deviation (PSD) was investigated using the corrected Akaike Information Criterion (AICc) from linear modeling. The best model to predict test-retest variability in the 24-2 VF included PSD, TFF, and FNs as dependent variables, while the best model for the 10-2 VF included PSD and average frequency of eye movements between 3° and 5° (P < 0.05 for all coefficients). Gaze tracking parameters are closely related to the reproducibility of VF results, and it would be beneficial to objectively use these parameters when estimating the reliability of VF tests.