Table 3.

Oculomotor measures of TBI.

Measure	Short description	Details
SacPurs (A)	Saccadic pursuit Gain	The saccadic (catch-up) pursuit is the saccadic component of the tracking that complements the smooth part. It is calculated per epoch as the speed (the slope of the linear fit in the 1–1.6 s post-stimulus onset time window) of the difference between two traces: full tracking minus the smooth tracking part (after saccades are removed and data aligned, Figure 2a), transformed to gain (fraction) by dividing the stimulus speed (4.4 deg./s for H and V, 6.2 for diagonals). Lower saccadic pursuit implies a smoother tracking.
OccDev (B)	Occluder-induced deviation from unoccluded pursuit	Deviation from unoccluded tracking during occlusion (Experiment 2). Each tracking trace was baseline-normalized relative to time 0 and converted to an unsigned trace, allowing opposite movement directions to be combined. A common unoccluded reference trajectory was computed from Experiment 1, separately for horizontal and vertical components. The primary measure was based on the horizontal component of deviation, including horizontal and diagonal trials. For each occluded epoch, the signed deviation from the reference was averaged over the 0.4–1 s time window and then averaged across epochs within observer after outlier rejection (>2 SD) (see Methods: Occluder-induced deviation analysis).
1st Speed (C)	Initial track speed	The time course of total eye velocity (drift, pursuit, and saccades) was first calculated per epoch (See Methods). The maxima were then calculated in the window 250–350 ms post-stimulus and averaged across epochs per participant.
sacRT (D)	Initial sac. Latency (catch-up)	A measure of the saccade response time, calculated per epoch as the latency of the first saccade in a time window of 250–800 ms relative to the onset of the pursuit target. It measures the speed of the open-loop process, in a condition in which the target direction is unknown in advance.
Pupil (E)	Pupil dilation, constriction	Change in pupil size calculated per epoch as the average in a 200–750 ms time window, expressed in percentage relative to the baseline pupil size (−0.5 to 0 s time window).
IOLag (F)	Vergence instability, via interocular lag	The horizontal interocular lag between the leading and the other eye time course was first calculated per epoch, and then normalized to time zero (the stimulus onset). Then, the maxima in the 1,250–2000 post-stimulus interval were calculated and averaged across epochs.

Six oculomotor measures that reflect the quality of tracking in the two experiments are listed together with a description of their exact computation method. The measures were computed per epoch and then averaged across epochs per participant and condition with outlier rejection (see the Methods section).