Drift Diffusion Model of Animacy Categorization Task Can Detect Patients with Mild Cognitive Impairment and Mild Alzheimer's Disease

Hamed  Karimi; Haniye  Marefat; Mahdiye  Khanbagi; Alireza  Karami; Zahra  Vahabi

doi:10.18502/fbt.v8i1.5857

Hamed Karimi Department of Brain and Cognitive Sciences, Cell Science Research Center, Royan Institute for Stem Cell Biology and Technology, ACECR, Tehran, Iran
Haniye Marefat School of Cognitive Sciences, Institute for Research in Fundamental Sciences, Tehran, Iran
Mahdiye Khanbagi
Alireza Karami
Zahra Vahabi

DOI: https://doi.org/10.18502/fbt.v8i1.5857

Keywords: Alzheimer's Disease; Mild Cognitive Impairment; Drift Diffusion Model; Machine Learning; Decision Making.

Abstract

Purpose: The process of neurodegeneration in Alzheimer's Disease (AD) is irreversible using current therapeutics. An earlier diagnosis of the disease can lead to earlier interventions, which will help patients sustain their cognitive abilities for longer. Individuals within the early stages of AD, shown to have trouble making confident and sounds decisions. Here we proposed a computational approach to quantify the decision-making ability in patients with mild cognitive impairment and mild AD.

Materials and Methods: To study the quantified decision-making abilities at the early stages of the disease, we took advantage of a 2-Alternative Forced-Choice (2AFC) task. We applied the Drift Diffusion Model to determine whether the information accumulation process in a categorization task is altered in patients with mild cognitive impairment and mild AD. We implemented a classification model to detect cognitive impairment based on the Drift Diffusion Model's estimated parameters.

Results: The results show a significant correlation of the classification score with the standard pen-and-paper tests, suggesting that the quantified decision-making parameters are undergoing significant change in patients with cognitive impairment.

Conclusion: We confirmed that the decision-making ability deteriorates at the early stages of AD. We introduced a computational approach for measuring the decline in decision-making and used that measurement to distinguish patients from healthy individuals.