Deep Learning in Mild Cognitive Impairment Diagnosis using Eye Movements and Image Content in Visual Memory Tasks

2025-06-28

AI Summary

Amid the global surge in dementia incidence, early detection of mild cognitive impairment (MCI) remains challenging. To address this, we propose a deep learning–based diagnostic framework that jointly models eye-tracking trajectories, spatial attention heatmaps (700×700 pixels), and semantic content of visual stimuli. We innovatively adapt the VTNet architecture to integrate temporal scanpath dynamics, high-resolution spatial attention maps, and stimulus image semantics. A systematic ablation study evaluates the impact of image resolution and task parameters on model performance. Trained and validated on a small-sample, short-duration visual memory task dataset, our model achieves 68% sensitivity and 76% specificity—performance comparable to state-of-the-art Alzheimer’s disease–focused studies. These results demonstrate the feasibility and effectiveness of noninvasive, scalable MCI screening based on naturalistic eye movement behavior.

Technology Category

• Computer Vision: Multi-modal Vision
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

The global prevalence of dementia is projected to double by 2050, highlighting the urgent need for scalable diagnostic tools. This study utilizes digital cognitive tasks with eye-tracking data correlated with memory processes to distinguish between Healthy Controls (HC) and Mild Cognitive Impairment (MCI), a precursor to dementia. A deep learning model based on VTNet was trained using eye-tracking data from 44 participants (24 MCI, 20 HCs) who performed a visual memory task. The model utilizes both time series and spatial data derived from eye-tracking. It was modified to incorporate scan paths, heat maps, and image content. These modifications also enabled testing parameters such as image resolution and task performance, analyzing their impact on model performance. The best model, utilizing $700 imes700px$ resolution heatmaps, achieved 68% sensitivity and 76% specificity. Despite operating under more challenging conditions (e.g., smaller dataset size, shorter task duration, or a less standardized task), the model's performance is comparable to an Alzheimer's study using similar methods (70% sensitivity and 73% specificity). These findings contribute to the development of automated diagnostic tools for MCI. Future work should focus on refining the model and using a standardized long-term visual memory task.

Problem

• Diagnosing Mild Cognitive Impairment using eye-tracking data
• Developing deep learning models for scalable dementia detection
• Analyzing visual memory tasks to improve MCI diagnosis accuracy

Innovation

• Deep learning model using eye-tracking data
• Incorporates scan paths, heat maps, image content
• Achieves 68% sensitivity and 76% specificity