Diagnostic power of rapid multifocal objective perimetry in Alzheimer’s

Abstract

Purpose : To examine the diagnostic power of novel measures obtained from three forms of pupil-mediated multifocal objective perimetry.

Methods : We examined 31 persons with early-stage Alzheimer’s (PwA, 19 males), and 36 control subjects (15 males). The ages of the PwA and control groups were well matched at 78.2 ± 6.50 years (mean ± SD) and 76.1 ± 5.81 years respectively. All tests were performed with an objectiveFIELD Analyser (OFA) by Konan Medical USA. Three test variants all presented 12 stimuli per eye respecting the horizontal and vertical meridians, extending to ± 30 degrees of the central visual field. One stimulus, W12, provided pseudo-randomly presented bright stimuli on a dim background (10 cd/m²) over 82 s. The other two differed in stimulus duration but both presented dark stimuli on a bright background (80 cd/m²). The durations were 82 and 360 s, hence the names W12_D82 and W12_D360. Diagnostic power was assessed by area under receiver operating characteristic plots (AUROC) and standardised effect size as measured by Hedge’s g. The input measure was the mean of the data for the worst performing 3 regions/eye. Separate analyses were done by splitting all subjects into males, and females, and by younger and older ages (> median age of 77). To cope with the unusual response waveform shapes produced by the dark stimuli we used the absolute value of deviations from the median delay for a given protocol relative to controls.

Results : The Table summarises the results. When split by ages (Table A) the best AUROC was for W12 and younger subjects (81.9 ± 4.08%, mean ± SD). For Hedge’s g old and young subjects were > 1.15 for W12_D82. When split by sex the outcomes were similar. Overall, the longer duration of W12_D360 appeared to offer no benefit relative to the shorter W12_D82 test. We also examined combining the per region sensitivities and delays of W12 and W12_D82. There was no advantage of combining the delays of the two tests, but combined sensitivities took AUROCs from 60.6 ± 6.41% to 80.1 ± 4.71% (p < 0.024).

Conclusions : Diagnostic performance for the shorter duration light and dark stimuli was reasonable. The fact that combined sensitivity increased diagnostic power suggests that they may characterise different aspects of Alzheimer’s. More work needs to be done to characterise the responses to dark stimuli.

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

 

Summary of diagnostic power