Detection of normal and slow saccades using implicit piecewise polynomial approximation.

Area of Science:

• Ophthalmology
• Neurology
• Biomedical Engineering

Background:

• Saccades, rapid eye movements, provide insights into cognition and neurological health.
• Abnormally slow saccades can indicate neurological disorders, but current detection algorithms struggle with these slow movements.

Purpose of the Study:

• To develop and evaluate a novel algorithm for accurately detecting both normal and slow saccades in eye movement data.
• To address the limitations of conventional saccade detection algorithms in identifying slow saccades.

Main Methods:

• Developed a new algorithm modeling saccadic waveforms as piecewise-quadratic signals.
• Employed noise reduction via optimization and subsequent velocity thresholding for saccade detection.
• Validated the algorithm using simulated, healthy subject, and patient-generated saccade data.

Main Results:

• The proposed algorithm demonstrated superior accuracy in detecting both normal and slow saccades compared to 10 other existing algorithms.
• Successfully identified slow saccades, which are often missed by conventional methods.

Conclusions:

• The developed algorithm offers a more reliable tool for analyzing saccades, particularly slow ones, in eye-tracking data.
• This advancement has potential implications for the diagnosis and monitoring of neurological conditions through eye movement analysis.