Related Experiment Videos
Measuring saccade curvature: a curve-fitting approach.
Casimir J H Ludwig1, Iain D Gilchrist
1Department of Experimental Psychology, University of Bristol, Bristol, England. c.ludwig@bristol.ac.uk
Behavior Research Methods, Instruments, & Computers : a Journal of the Psychonomic Society, Inc
|February 5, 2003
Summary
Saccade curvature metrics were compared, revealing quadratic polynomial fits effectively capture single and double-curved saccades, offering a robust measure for motor program competition.
Area of Science:
- Oculomotor research
- Computational neuroscience
- Motor control
Background:
- Saccade curvature quantifies competing motor programs.
- Various metrics exist, but their relationships are unclear.
- Novel curve-fitting methods offer potential improvements.
Purpose of the Study:
- Compare existing saccade curvature metrics.
- Evaluate novel curve-fitting approaches.
- Identify the most robust and informative metric.
Main Methods:
- Compared initial deviation, maximum curvature, and area-based metrics.
- Developed and applied second- and third-order polynomial curve fits.
- Analyzed saccade trajectories for single and double curvature.
Main Results:
- Initial deviation metrics showed weak correlation with maximum curvature.
- Maximum curvature strongly correlated with area-based and polynomial fits.
- Identified a subset of double-curved saccades using curve fitting.
Conclusions:
- Quadratic polynomial fits effectively measure saccade curvature in both single and double-curved trajectories.
- Curve-fitting methods, especially quadratic fits, are less susceptible to sampling noise.
- This approach provides a more reliable quantification of saccade curvature for motor program analysis.