Computational modeling of selective attention differentiates subtypes of amnestic mild cognitive impairment.

Abstract

Individuals with amnestic mild cognitive impairment (aMCI), a prodromal stage of Alzheimer's disease and other dementias, show inhibition deficits in addition to episodic memory. How the latent processes of selective attention (i.e., from perception to motor response) contribute to these inhibition deficits remains unclear. Therefore, the present study examined contributions of selective attention to aMCI-related inhibition deficits using computational modeling of attentional dynamics. Two models of selective attention - the dual-stage two-phase model and the shrinking spotlight model - were fitted to individual participant data from a flanker task completed by 34 individuals with single-domain aMCI (sdaMCI, 66-86 years), 20 individuals with multiple-domain aMCI (mdaMCI, 68-88 years), and 52 healthy controls (64-88 years). Findings showed greater commission errors in the mdaMCI group compared to controls. Final-fitting model parameters indicated inhibitory and early perceptual deficits in mdaMCI , and impaired spatial allocation of attention in both MCI groups. Model parameters differentiated mdaMCI from sdaMCI and controls with moderate-to-high sensitivity and specificity. Impairments in perception and selective attention may contribute to inhibition deficits in both aMCI subtypes.