Abstract
Purpose :
We have previously shown [1] using computer simulation that the Ψ algorithm [2] could be used in a clinical setting for determining whether a Frequency of Seeing curve (FoS) is steep or flat. Here we investigate the performance of Ψ in patients with moderate to advanced glaucoma.
Methods :
19 patients (MD mean -5.08 dB range [-14.72, -1.72]) were tested in 3 locations of their visual field with Ψ during a visit for the IPPOS clinical trial [3]. The locations were chosen by author VT to be spatially separated and in previously measured normal areas of the VF. Ψ used a domain of thresholds 20 to 40 dB, slopes of 1 to 5 dB (standard deviation of Cumulative Gaussian) and 5 possible values for the asymptotes of the FoS curve ranging from 0 to 20%. The procedure ran for 27 presentations at each location using the OPI [4] interface to the Compass perimeter (CenterVue, Italy). Patients also underwent a standard 24-2 VF test on the Compass perimeter (Zest Faster). Locations that measured close to the floor of the range in Ψ (<22 dB) were discarded from the analysis.
Threshold measurements were compared between Ψ and Zest Faster. Slopes are reported and compared with thresholds and with previous models of the slope-threshold relationship [5].
Results :
34 locations in 17 eyes were analysed. The mean difference between the Ψ and Zest Faster thresholds was -0.76 dB (sd 3.14, mean-abs 2.24, range [0.06, 9.00]). A least-squares log-linear fit shows slope = exp(2.55 - 0.061 * threshold) (both parameters t-test p < 0.02). This is similar to the Henson et al formula [5] of slope = exp(3.27 - 0.081 * threshold). While on average the slopes followed Henson et al, there was substantial deviation from the average for any individual. Steep slopes (near 1.0) were present across the range of thresholds from 25 dB to 36 dB. 5 of the 17 eyes had all slopes below 2.
Conclusions :
The average relationship between threshold and slope measured clinically using Ψ follows known patterns. However, 5 of the 17 eyes showed steep slopes well below population average [5]. These eyes are candidates for the application of tighter criteria for calling glaucomatous progression, allowing earlier detection of progression if progression is present [6].
[1] Turpin et al. 2010 Vis Res 50
[2] Kontsevich et al. 1999 Vis Res 39
[3] IPPOS Trial. https://tinyurl.com/mwchf89j
[4] Turpin et al. 2012 J Vis 12
[5] Henson et al. 2000 IOVS 41
[6] Turpin et al. 2005 Vis Res 45
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.