Fig 1.
Comparison of data recorded from the analog versus the digital output of the video tracker.
A two second epoch of analog versus digital eye movement recordings for each subject (indicated at the right) for the horizontal (left) and vertical (right) eye positions. Red traces are the digital signals obtained from the analog output of the video tracker as described in the Methods section. Pink traces are the corresponding data obtained from the digital output of the video tracker. Both signals are nearly identical.
Fig 2.
Microsaccade and overshoot detection.
Here we show the horizontal (top black trace) and vertical (bottom black trace) EyeLink data from subject 5 over a 2.5 second epoch of time. Microsaccades were manually detected by aligning (using the mouse as control) the cross hairs, shown simultaneously on the horizontal and vertical axes, with the start of the microsaccade and clicking the mouse, and then aligning to the stop of the microsaccade and clicking the mouse. Overshoots were detected in the same way. An overshoot always began at the sample after the end of the microsaccade. The red crosses on the black traces indicate the defining samples of microsaccades and the blue crosses indicate the defining samples of the overshoots.
Fig 3.
Coherence of detection between manually detected microsaccades and microsaccades detected with an objective algorithm, for the search coil (top row) and video tracker (bottom row) data.
(A) Percentage of search coil microsaccades detected manually that were also detected with the objective algorithm. (B) Percentage of search coil microsaccades detected manually that were also detected with the objective algorithm. (A) Percentage of video tracker microsaccades detected manually that were also detected with the objective algorithm. (B) Percentage of video tracker microsaccades detected manually that were also detected with the objective algorithm. Error bars represent the S.E.M. across participants (n = 4).
Fig 4.
Examples of simultaneous horizontal and vertical eye position traces obtained with the search coil and video tracking, for each participant.
Red triangles represent microsaccades detected with the video tracker; blue triangles represent microsaccades detected with the search coil. Note that microsaccades were detected by viewing the vertical and horizontal traces simultaneously. For illustration purposes, a small offset of 0.2 deg has been added to the search coil data.
Fig 5.
Horizontal components from the vectors used to calculate microsaccade magnitude, displacement, and overshoot size (see Methods for the precise definitions of these terms).
The vertical components of the vectors were defined similarly. Arrows indicate the magnitude, displacement, and overshoot of a video-tracked microsaccade. For illustration purposes, a small offset of 0.2 deg has been added to the search coil data.
Fig 6.
Coherence of detection between the two recording systems.
(A) Percentage of microsaccades detected with the search coil that were also detected with the video tracker. (B) Percentage of microsaccades detected with the video tracker that were also detected with the search coil. Error bars represent the S.E.M. across participants (n = 4).
Fig 7.
Magnitude distributions for microsaccades detected in only one recording system or in both systems.
Magnitude distributions of microsaccades detected in only one system (A–B) or both systems (C–D). Detection system indicated at the top of each panel.
Fig 8.
Comparison of microsaccade properties for microsaccades detected in both recording systems.
Scatter plots for microsaccade magnitude (A), displacement (B), overshoot size (C), peak velocity (D), mean velocity (E), and duration (F). Each dot in (A–F) represents one microsaccade, with the property as measured with the video tracker on the y-axis and with the search coil on the x-axis. Average microsaccade magnitude (G), displacement (H), overshoot size (I), peak velocity (J), mean velocity (K), and duration (L), for each participant (i.e. average data from panels A–F). Each dot represents the average microsaccade property for one participant as measured with the video tracker on the y-axis and with the search coil on the x-axis. In the legend in panel (A), N indicates the number of common microsaccades detected in both data sets. Bar insets in (G–L) are the video tracker (red) and search coil (blue) averages across participants. Error bars represent the S.E.M. across participants (n = 4). * indicates statistical significance with a p-value < 0.05 (two-tailed paired t-tests).
Fig 9.
Examples of simultaneous horizontal drift traces (green lines) obtained with the search coil and video tracker, for each participant contributing to the drift analyses.
For illustration purposes, a small offset of 0.2 deg has been added to the search coil data.
Table 1.
Drift descriptive statistics and correlations between ocular drift measured with the video tracker and ocular drift measured with the search coil.
Table 2.
Drift and pupil size.
Fig 10.
Distribution of pupil size deviations from the mean pupil size for all subjects considered in the drift analyses (n = 3).