Purchase this article with an account.
Ali Hashemi, Eugenie Roudaia, Nicole D. Anderson, Claude Alain, Rosanne Aleong, Nasreen Khatri, Morris Freedman, Allison B. Sekuler; Behavioural and electrophysiological measures of visual processing for early detection of Alzheimer’s disease. Journal of Vision 2020;20(11):1624. doi: https://doi.org/10.1167/jov.20.11.1624.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Alzheimer’s disease (AD) begins years before clinical diagnosis, but there are no simple, cost-effective methods to identify individuals in preclinical stages of AD, when interventions are most likely to succeed. Individuals with AD show deficits in multiple visual functions thought to reflect changes in parieto-occipital and temporal brain regions. However, we know little about how vision changes during preclinical AD – a critical step in determining whether visual tasks can predict AD. Here, we present psychophysical and electrophysiological results for two simplified tasks collected from individuals with mild cognitive impairment (MCI; N=8; Age=61-88, MoCA=20-26) and normal cognition (NC; N=8; Age=62-82, MoCA=23-30). Methods: Face Identification: participants selected which of two briefly presented faces matched a target face identity, measuring accuracy and response time. Contour Integration: we measured density thresholds to identify the global orientation of a spiral contour embedded in a field of cluttering elements. In both tasks, event-related potentials (ERPs) were acquired using the consumer-focused Muse system. Results: Face Identification: the MCI group showed slightly, but not significantly, worse accuracy (Mdiff=0.06, 95%CI=[-0.06,0.19]) and slower response times (Mdiff=-0.35, 95%CI=[-0.96,0.26]) than the NC group. MCI N170s were reduced in amplitude (Mdiff=-1.42𝜇V, 95%CI=[-3.96,1.12]) and delayed (Mdiff=-15.1ms, 95%CI=[-33.8,3.56]) relative to NCs, although these differences also were not significant. Contour Integration: In contrast to the results from face identification, there was a large group difference in both density thresholds (Mdiff=0.45, 95%CI=[0.24,3.56]) and N1 latency (Mdiff=-40.5ms, 95%CI=[-67.7,-13.4]), but not in N1 amplitude (Mdiff=-0.29𝜇V, 95%CI=[-2.84, 2.27]). Delayed N1s were significantly correlated with worse density thresholds (r=-0.78) and lower MoCA scores (r=-0.62), and lower MoCA scores correlated with worse density thresholds (r=0.69). These are the first results showing that behavioural and ERP measures of contour perception may distinguish between normal cognition and MCI. Thus, simple visual tasks may provide viable candidates for early markers of preclinical AD.
This PDF is available to Subscribers Only