Eye tracking based detection of mild cognitive impairment: A review
Hasnain Ali Shah, Salman Khalil, Sami Andberg, Anne M. Koivisto, Roman Bednarik
Abstract
Mild Cognitive Impairment (MCI) is a transitional stage between normal age-related cognitive decline and the more severe cognitive loss seen in dementia. MCI presents a higher risk of progressing to Alzheimer’s and other neurodegenerative conditions despite its early symptoms overlapping with normal cognitive aging. Early identification is crucial for timely intervention, as no cure exists, and current treatments only slow disease progression. Eye tracking is a promising diagnostic solution, offering a non-invasive method to examine cognitive functions related to eye movements. In this article, we comprehensively reviewed how recent research studies have implemented eye-tracking technology to study cognitive disorders, focusing on MCI. We reviewed a broad spectrum of research studies employing eye-tracking methods and provided a cohesive overview of the different methodological approaches. We highlighted key advancements in detecting MCI using commonly analyzed eye movements and innovative eye-tracking techniques. Additionally, we analyzed recent studies that applied statistical, machine learning, and deep learning methods to eye-tracking data for MCI detection, followed by a detailed discussion on their challenges and limitations towards clinical applications. • Comprehensive review of eye-tracking in cognitive disorder research, focusing on MCI/AD. • Survey of gaze-based biomarkers and emerging paradigms in early disease detection. • Analysis of statistical, ML, and deep learning methods for diagnostic classification. • Discussion of clinical challenges in feasibility, generalizability, and deployment.