An automated segmentation approach to calibrating infantile nystagmus waveforms

Area of Science:

• Ophthalmology
• Neuroscience
• Biomedical Engineering

Background:

• Infantile nystagmus (IN) involves repetitive eye movements with a foveation period crucial for best vision.
• Accurate eye-tracking calibration is challenging due to continuous IN eye movements.
• Manual selection of foveation periods for calibration is time-consuming and impacts waveform parameter determination.

Purpose of the Study:

• To develop and validate an automated method for segmenting infantile nystagmus waveforms.
• To precisely determine foveation positions for calibrating eye-tracking systems.
• To facilitate rapid clinical quantification and enable gaze-contingent research in IN patients.

Main Methods:

• Development of an automated algorithm for segmenting infantile nystagmus waveforms.
• Utilizing the segmented waveforms to identify foveation positions.
• Comparison of automated calibration results with expert manual marking.

Main Results:

• The automated method achieved an average "point of regard" accuracy within 0.21° of expert manual identification.
• Demonstrated the ability to rapidly quantify IN waveforms.
• Validated the potential for improved eye-tracker calibration in IN.

Conclusions:

• The automated method provides a fast and accurate approach to segmenting IN waveforms and identifying foveation positions.
• This technique significantly enhances the calibration of eye-tracking systems for individuals with infantile nystagmus.
• Enables more efficient clinical assessment and opens possibilities for advanced gaze-contingent research in IN.