EM-COGLOAD: An investigation into age and cognitive load detection using eye tracking and deep learning

Area of Science:

• Neuroscience
• Biomarkers
• Artificial Intelligence

Background:

• Alzheimer's Disease (AD) is a leading cause of elderly disability.
• Eye movement behavior is a potential non-invasive biomarker for early AD detection.
• Changes in eye movement are detectable early in the disease progression.

Purpose of the Study:

• Introduce the EM-COGLOAD dataset for eye movement and cognitive load research.
• Explore deep learning techniques for analyzing eye movement data.
• Assess the potential of eye movement as a biomarker for cognitive decline and aging.

Main Methods:

• Collected eye movement data from 75 healthy adults using a dual-task paradigm.
• Induced varying cognitive loads during visual tracking tasks.
• Applied time series classification with deep learning models to eye movement traces.

Main Results:

• Convolutional neural networks achieved 87.5% accuracy distinguishing low vs. high cognitive load.
• Convolutional neural networks achieved 76% accuracy distinguishing oldest vs. youngest age groups.
• Demonstrated feasibility of using deep learning on eye movement data for cognitive assessment.

Conclusions:

• Eye movement patterns under cognitive load can be effectively classified using deep learning.
• Eye tracking shows potential for early detection of cognitive changes associated with aging and potentially Alzheimer's Disease.
• The EM-COGLOAD dataset provides a valuable resource for further research in this area.