Sampling Rate Guidelines for Eye-Tracking in Neurological Disorders: A Task-Specific Multimetric Analysis.

Area of Science:

• Ophthalmology
• Neuroscience
• Biomedical Engineering

Background:

• Eye-tracking is crucial for detecting neurodegenerative disorders (ND).
• Inconsistent sampling rates across devices challenge data quality and feature integrity.
• Precise eye movement measurement is essential for identifying disease-specific patterns.

Purpose of the Study:

• To evaluate the impact of downsampling on eye-tracking data quality and feature integrity.
• To determine task-specific sampling rate requirements for accurate neurodegenerative disorder research.
• To propose metrics for systematically assessing data degradation due to reduced sampling rates.

Main Methods:

• Three metrics (TOST Test Equivalence, Mean Absolute Error Percentage, Kullback-Leibler Divergence) were used to assess data quality.
• Eye-tracking data from prosaccade, text reading, and picture description tasks were analyzed.
• The effects of downsampling on saccadic dynamics, fixation features, and overall data fidelity were evaluated.

Main Results:

• Task-specific sampling rates are critical: prosaccade tasks require ≥400Hz, while reading and picture description tasks need ≥600Hz.
• Downsampling significantly degrades key features used to differentiate neurological disorders.
• A minimum of ≥200Hz is necessary to preserve the discriminative power of saccade and fixation features.

Conclusions:

• Standardizing eye-tracking sampling rates is vital for research reproducibility and biomarker validation in neurodegenerative diseases.
• Task- and feature-specific recommendations for sampling rates can balance analytical rigor with practical constraints.
• These findings offer guidelines for reliable eye-tracking protocols in Alzheimer's, Parkinson's, and related disorders.