Signal/noise ratio for area-modulated perimetric stimuli optimised to measure changes in spatial and temporal summation in glaucoma

Abstract

Purpose : To compare novel area-modulated stimuli, designed to measure two functional biomarkers in glaucoma (spatial and temporal summation) with conventional luminance-modulated stimuli.

Methods : Disease signal, response variability, and signal/noise ratio (SNR) were determined for two area-modulated, fixed-luminance stimuli of 200ms (AMS200 [ΔI: 5cd/m2]) or 16ms (AMS16 [ΔI: 9cd/m2]) duration, and two luminance-modulated, fixed-area stimuli of 200ms duration (Goldmann III [GIII] and Goldmann V [GV]) at 4 locations (9.9°-20°eccentricity) in 52 glaucoma patients (median [IQR] age (years): 69.1 [58.9, 75.3], MD (dB): -3.1 [-1.3, -6.8]) and 53 controls (age: 61.7 [51.9, 69.7], MD: 0.5 [-0.7, 1.5]). Stimulus levels were defined by a common energy scale (luminance × area × duration) and equated across all stimuli. Threshold (50% seen) and response variability (slope, ‘noise’) were derived from psychometric functions following a 3-step sampling process. Total deviation (TD, ‘disease signal’) was calculated as the difference between measured threshold and that predicted for an age-matched control. SNR was calculated as TD/response variability. The effect of stimulus form on SNR was analysed with a linear mixed effects model.

Results : AMS16 had the highest disease signal with a mean difference in disease signal from that of GIII (ΔDS) of 0.28 log Energy (Fig.1A). AMS200 had the next highest disease signal (ΔDS: 0.15 log Energy), and GV had the smallest (ΔDS: -0.19 log Energy). Response variability was lowest for GV in early damage, but more uniform with depth of defect for area-modulated than for luminance-modulated stimuli (Fig.1B). Compared to GIII, SNR was significantly higher for AMS16 (1.59±0.40, p<0.0001) and AMS200 (0.52±0.23, p=0.02), and lower (though not statistically significant) for GV (-0.33±0.23, p=0.15); (Fig.1C).

Conclusions : Area-modulated stimuli, designed to measure changes in spatial and temporal summation, have a higher disease signal, more uniform noise, and higher SNR than luminance-modulated perimetric stimuli. The AMS16, which exploits altered temporal summation, yields additional SNR compared to the AMS200. Despite low response variability, GV had the smallest SNR.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.

 

Fig.1: For all stimuli and pooled across locations, (A) Disease signal (TD) as a function of GIII TD; (B) Response variability as a function of stimulus-specific TD; (C) SNR.