Purpose: Smooth pursuit (SP) eye movements are used to track moving targets. If target velocity exceeds ~60deg/s or the target is unpredictable, high velocity catch-up saccades (CS) occur. To study SP eye movements or CS separately, a method is needed to differentiate them. Here we demonstrate a novel objective criterion for marking CS utilizing the derivative of ocular acceleration (dA).

Methods: Search coils were used to monitor eye position during tracking of an unpredictable 2D pursuit target (0.24-1.25Hz). Horizontal and vertical components were analyzed separately. Automated software was developed to determine the start and end of CS. The start of the CS was the peak dA point prior to the peak velocity point of the CS, and the end was the minimum dA point following the peak velocity point. One experienced author visually marked the start and end of 21 horizontal and 26 vertical saccades from the same eye trace.

Results: Comparisons (paired t-test) were made between CS start and end times as determined with the automated method and visual inspection. Small but significant differences occurred for start (mean difference = 7.1msec, p<0.0005) and end (mean difference = 6.7msec, p<0.0005) times.

Conclusion: The automated method is an accurate, rapid method for determining the start and end of CS. Because start times were later and end times earlier with visual inspection, the small differences between the automated and visual inspection methods are probably due to difficulties in visually detecting the subtle changes in acceleration at the beginning and end of CS.