Real-Time Biometric Analysis of Eye Movement Patterns in Human Spatial Navigation

Human spatial navigation is crucial for daily functioning and cognitive development. It mainly relies on the egocentric and allocentric systems with different developmental trajectories. This study presents a data-driven approach to quantify spatial navigation performance through eye movement pattern analysis. Using the Leiden Navigation Test (LNT) as our experimental framework, we extracted and analyzed multiple eye-tracking metrics to identify distinct visual signatures associated with different navigation strategies. Our correlation analysis revealed significant patterns: landmark recognition tasks demonstrated correlations with blink amplitude and left saccade velocity <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(r=0.38-0.4,\ p\leq 0.05)$</tex>, while allocentric spatial processing strongly correlated with vertical scanning parameters <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(r=0.41-0.5,\ p &lt; 0.05)$</tex>. Most notably, path survey tasks showed robust correlations with left saccade metrics <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(r=0.62-0.68, p &lt; 0.0001)$</tex> and horizontal eye movement energy <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(r=0.54,\ p &lt; 0.01)$</tex>. Path route tasks exhibited inverse relationships with fixation frequency and saccade duration <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(r=-0.43$</tex> to -0.52, <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$p &lt; 0.02)$</tex>. These findings establish a quantitative foundation for developing automated assessment tools and potential real-time navigation assistance systems based on eye movement patterns.