The relationship of visual threshold and reaction time to visual field eccentricity with conventional automated perimetry

To investigate the relationship between reaction time (RT), stimulus intensity and visual field eccentricity. We generated frequency of seeing (FOS) curves and measured RTs by testing 10 perimetrically experienced normal subjects with a Humphrey perimeter controlled by a custom program. Subjects were tested from 10° to 50° eccentricity along the nasal horizontal meridian in 10° increments. A range of 20 dB, centered on threshold, was tested in 1 dB steps along with 60 and 0 dB intensities as catch trials. Twenty repetitions for each intensity at each location were used. Linear regression showed a significant increase in suprathreshold RT (to the 0 dB stimulus) with increasing eccentricity. The RT at the calculated FOS 50% threshold was prolonged by about 200 ms compared with the RT using the 0 dB target at the equivalent eccentricities. Also, when the difference between the RT at 0 dB stimulus and the RT at threshold was regressed against visual field eccentricity there was a significant decrease with eccentricity. When the RT_pi (RT prolongation from threshold relative to the 0 dB stimulus) was plotted as a function of decreasing stimulus attenuation, the results fit the function RT_pi = a + bi³ (i = stimulus intensity) with r² > 0.94 at all eccentricities. However, the slope of the function flattened with increasing eccentricity. Using conventional automated perimetry stimuli in perimetrically experienced young subjects, suprathreshold RT increases but threshold RT prolongation decreases with increasing visual field eccentricity. RT fits a power function with decreasing stimulus attenuation but the slope flattens with eccentricity. This relationship found along the nasal horizontal meridian may allow use of RT to cross-check threshold results or to define response windows for reliability indices of conventional automated perimetry.