Purchase this article with an account.
Alyssa C Ehrlich, Matthew Nguyen, Ali S Raza, Ieva Sliesoraityte, Andrea Mast, Ulrich Schiefer, Donald C Hood; Using High-Density Perimetry to Test a Model of Glaucomatous Damage of the Macula and to Assess the Placement of Visual Field Test Points. Invest. Ophthalmol. Vis. Sci. 2014;55(13):5644.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To use high-density perimetry to test a model  of local glaucomatous damage to the macula and to assess the placement of visual field (VF) test points used to detect this damage.
One eye from each of 31 patients with a glaucomatous arcuate VF defect in the central 10° of the upper hemifield was tested using a custom VF (Octopus, Haag-Streit, Inc.) with a rectangular 1°x1° test grid (double the test point density of the 10-2). Individual plots of total deviation (TD) from age-matched control values were generated, with point locations morphed as per displacement of retinal ganglion cells near the fovea. A model, which predicts a “macular vulnerability zone” (MVZ, magenta in Fig. 1) and a less vulnerable region (cyan), was superimposed on each VF plot and aligned with patients’ fovea and disc centers by scaling and rotating. The VF data were also used to simulate a 6° grid (like the 24-2) and the Octopus G program, as well as a 6° grid with 2 additional points. Three different pairs of points were chosen: (1) from the MVZ; (2) from the less vulnerable region; (3) by a MATLAB program that maximized the average number of abnormal points (TD≤-5 dB). All tests and simulations were restricted to the central ±10°.
The average percent of VF points with TD≤-5 dB was significantly greater for the MVZ, 61.2% compared to 15.8% for the less vulnerable region (p<0.001; paired t-test). After scaling and rotating the model, these percentages were essentially the same, 61.7% and 15.3% (p=0.62, p=0.42). Moreover, the average number of abnormal points (see Table 1) was greater for the G program (3.94) than for the 6° grid (3.06). Yet this value was increased further by adding 2 points in the MVZ to the 6° grid (4.61), a value close to that for the pattern derived by empirical optimization (4.65).
The spatial pattern of high-density VF loss was consistent with the model, and scaling and rotating the model to align it with fovea and disc centers had little effect on this agreement. VF simulations suggest that the detection of glaucomatous damage to the macula could be improved by adding 2 test points in the MVZ to 6° grid test patterns (e.g. the 24-2). 1. Hood et al, PRER, 2013; 2. Raza et al, AO, 2011; 3. Jansonius et al, Vis Res, 2009.
This PDF is available to Subscribers Only