Detecting mild cognitive impairment by quantifying oculometric features during variable cognitive workload: The novel rapid Oculo‐Cognitive Addition Test

Abstract

AbstractBackgroundThe need exists for a device that can rapidly and accurately assess mild cognitive impairment (MCI). Because the link between oculometrics and cognition is well‐documented, the Oculo‐Cognitive Addition Test (OCAT) was developed which tracks users’ eye movements as they complete a simple mental addition test.MethodOCAT consisted of 12 trials of summing three consecutive single‐digit numbers, shown separately on three consecutive screens (Figure 1). The first, second, and third screen in each trial corresponded to low, medium, and high cognitive workload. Using an infinity‐loop pattern of 24 symmetrical positions, eye movements were captured systematically in horizontal, vertical, and diagonal directions (Figure 2).MCI was simulated by exposing subjects to 5 minutes of hypoxic gas mixture (8% O2 + balance N2). 20 subjects completed OCAT in baseline (i.e. normal room air), hypoxic, and post‐baseline (i.e. 20 minutes of room air) conditions. Raw scan‐paths were collected with an eye‐tracking device. Oculometrics pertaining to fixations and saccades were extracted. Statistical significance was assessed using repeated measures ANOVA. Effect size and clinical significance was evaluated using computed Cohen’s d.ResultOCAT completion time significantly increased from baseline to hypoxia, which corresponded to a significant decrease in cerebral oxygenation from baseline to hypoxia (Figure 3). From baseline to hypoxia, saccadic latency significantly increased while saccadic velocity significantly decreased in the diagonal direction, however, results in the horizontal and vertical directions were not significant (Figure 4). Fixation time significantly increased from baseline to hypoxia in medium and high cognitive workload, but did not significantly change in low cognitive workload (Figure 5). From baseline to hypoxia, fixation size did not significantly change in any cognitive workload, however, the fluctuation in fixation size significantly increased in medium cognitive workload. From baseline to hypoxia, fixation area significantly increased in low and high cognitive workload, but did not significantly change in medium cognitive workload.ConclusionOCAT can detect subtle changes in oculometrics that correlate with MCI. This study provides rationale for providing varying degrees of cognitive load to elucidate subtle changes in oculometrics that may otherwise go undetected in neurocognitive tests that do not increase cognitive load.