Reconstruction of eye movements during blinks

Area of Science:

• Ophthalmology
• Cognitive Science
• Data Science

Background:

• Eye movement data is crucial for validating reading research.
• Blinks cause significant data loss during reading, particularly for vulnerable groups.
• Existing data reconstruction methods are often insufficient.

Purpose of the Study:

• To present a novel method for reconstructing horizontal eye position during blinks.
• To address data loss issues in eye movement research during reading.

Main Methods:

• Utilizing symbolic sequence dynamics of eye movements.
• Leveraging pre- and post-blink eye movement data for reconstruction.
• Comparing the new method against simpler interpolation techniques.

Main Results:

• The proposed reconstruction method significantly outperforms simpler interpolation approaches.
• Reconstructed data shows no significant deviation from typical reader behavior.
• The method effectively recovers horizontal eye position during blinks.

Conclusions:

• This method enhances the reliability of eye movement data in reading research.
• It offers a superior solution for mitigating blink-induced data loss.
• The technique supports more accurate analysis for all reading populations.