Digital Marker for Early Screening of Mild Cognitive Impairment Through Hand and Eye Movement Analysis in Virtual...

Area of Science:

• Neurology
• Digital Health
• Biomarkers

Background:

• Alzheimer disease (AD) is rising globally, making early detection of mild cognitive impairment (MCI), a precursor to AD, crucial.
• Current biomarkers like CSF amyloid and MRI are costly and invasive.
• Digital markers analyzing daily behavioral data offer a novel, accessible alternative for cognitive assessment.

Purpose of the Study:

• To identify key behavioral features from a virtual kiosk test that can significantly differentiate MCI patients from healthy controls.
• To develop a machine learning model for early MCI screening using these identified behavioral features.

Main Methods:

• Recruited 51 participants (31 MCI, 20 healthy controls).
• Developed and administered a virtual kiosk test recording hand and eye movements.
• Computed four behavioral features: hand movement speed, fixation duration proportion, time to completion, and number of errors.
• Utilized t-tests for group comparisons and a support vector machine for machine learning model development.

Main Results:

• All four behavioral features showed statistically significant differences between MCI patients and healthy controls.
• MCI patients exhibited slower hand speed, lower fixation duration, longer completion times, and more errors.
• The machine learning model achieved 93.3% accuracy, 100% sensitivity, 83.3% specificity, 90% precision, and 94.7% F1-score.

Conclusions:

• The virtual kiosk test, analyzing hand and eye movements, shows potential as a digital marker for early MCI screening.
• This VR-based digital marker is cost-effective, collects ecologically valid data quickly (5-15 min), and is suitable for early MCI detection.
• Further research is needed to confirm the reliability and validity of this digital screening approach.