Abstract
Purpose :
Visual field (VF) examination by standard automated perimetry (SAP) is an important but challenging, or sometimes impractical, clinical assessment. We propose a new approach to central VF assessment using an inexpensive eye tracker and implement it on a tablet computer (Eyecatcher). We examine the usability of Eyecatcher and concordance with results from SAP in people with glaucoma and visually healthy peers.
Methods :
Eyecatcher utilises novel software producing a sequence of stimuli (spot size of 0.43°) of varying contrast presented at different locations on a tablet computer with an attached ‘clip-on’ eye tracker (Tobii EyeX; Tobii, Danderyd, Sweden). The test task is to simply follow the sequence of stimuli without any button pressing. Testing is done monocularly without a chin rest and instructions are simply, “move your eye to keep looking at the spot”. Start and end fixations are then processed in a novel fashion to determine stimuli that are seen and unseen to build a continuous map of sensitivity loss across a VF of approximately 20°. A group of glaucoma patients and visually healthy peers were tested on Eyecatcher and SAP (Humphrey Field Analyser [HFA] SITA standard 24-2); all were experienced tests takers with SAP. The VF surface generated from Eyecatcher was thesholded to give suprathreshold defects at SAP locations; these were then compared to SAP defects defined as sensitivity below 25dB. Percentage of concordant ‘defect’ and ‘no defect’ locations per eye was used as an outcome measure for agreement between the VFs. In addition participants were asked to rate preferences for each examination on a 5-point Likert scale.
Results :
Twenty-two eyes from 12 patients (median [interquartile range; IQR] age 70 [67 to 76] years) and 12 eyes from 6 visually healthy people (76 [73 to 79] years) were examined. Eyes with glaucoma had a range of VF loss; median HFA MD -12 [-17 to -6]dB. Median suprathreshold defect concordance between Eyecatcher and SAP was 83 [62 to 90]% and 96 [92 to 97]% in patients and visually healthy people respectively (Fig 1). Participants preferred Eyecatcher to SAP (median 4/5 vs 3/5; p<0.01 Wilcoxon test).
Conclusions :
Visual fields measured by Eyecatcher show good concordance with those measured by SAP. Eyecatcher, requiring very simple user interaction, was a preferred examination in this group of people. Perimetry using an inexpensive eye tracker on a tablet computer is feasible.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.