Effects of binocularity and eye dominance on visually-driven ocular tracking

Abstract

Introduction: We used 18 oculomotor performance metrics (oculometrics) to capture largely independent features of human ocular tracking. Our primary goal was to examine tracking eye movements in a healthy population under monocular and binocular viewing, as well as to examine the potential effects of line-of-sight eye dominance and spatial/directional tuning.

Methods: We compared the ocular responses of 17 healthy well-rested participants using a radial step-ramp paradigm under three viewing conditions: both-eyes viewing, left-eye viewing, and right-eye viewing.

Results: Our findings revealed that binocular viewing enhanced performance over that during monocular viewing for 11 oculometrics, with eye dominance associated with the selective enhancement of 3 oculometrics of visual motion processing. A comparison of binocular and dominant-eye viewing allowed us to segregate the direct enhancements of binocularity per se from those due simply to the inclusion of the dominant eye in binocular viewing and showed that viewing with two eyes is only directly responsible for the enhancement of 9 oculometrics. Our examination of spatial/directional tuning revealed largely isotropic enhancement due to binocularity, as well as several anisotropies in retinal functional processing: (1) a Nasal-Temporal asymmetry for pursuit latency and direction noise, and a Superior-Inferior asymmetry for latency, and (2) anisotropic enhancement in initial acceleration and direction noise (primarily for nasal retina) and speed noise (primarily for superior retina) when viewing through the dominant eye. We also documented Horizontal-Vertical anisotropies in initial acceleration, steady-state gain, proportion smooth, and speed responsiveness for both monocular and binocular viewing.

Conclusion: Our findings demonstrate that there is isotropic enhancement from binocular viewing per se across a wide range of visuomotor features and that important normative characteristics of visual motion processing are shaped by retinal processing non-uniformly across visual space, modulated by eye dominance and perhaps related to previously found normative structural anisotropies in retinal thickness. This constellation of findings characterizes the subtle natural non-linear variations in visuomotor performance to provide insight into the relative roles of the retina and other brain areas in shaping visuomotor performance and to enable the detection of neurological and ophthalmological impairment through comparison with properly matched baselines in support of future research and clinical applications.

Keywords: binocular; monocular; oculometric; saccades; smooth pursuit.

Figure 1

Typical raw eye-velocity traces from individual trials under three different viewing conditions for the same healthy participant. The top trace is the response to 22 deg/s ramp target motion 24 deg upward from pure leftward motion under binocular viewing. The middle trace is the response to 18 deg/s ramp target motion 24 deg downward from pure leftward motion under left (dominant) eye viewing. The bottom trace is the response to 20 deg/s ramp target motion 28 deg downward from pure rightward motion under right (non-dominant) eye viewing. Note that, while all three traces are qualitatively similar, they differ in important details. The binocular trace has a shorter pursuit latency compared to that in either monocular viewing condition. The two monocular conditions have similar pursuit latencies. Note also that the binocular and dominant eye responses have similarly brisk accelerations, while that for the non-dominant response is more sluggish. Lastly, the saccade amplitude (numbers adjacent to the saccadic velocity spike) for binocular viewing was slightly smaller compared to either monocular viewing condition. While the distributions of oculometric responses both within and across participants in these three conditions overlap considerably, the results in this figure illustrate the average behavior across our population. For examples of larger qualitative differences in oculometrics due to a behavioral stressor (see Figures 5A,B of Tyson et al., 2021).

Figure 2

Plots of oculometric values under binocular vs. monocular viewing for (A) Latency, (B) Initial Acceleration, (C) Direction Noise, (D) Direction Asymmetry, (E) Saccade Rate, (F) Saccade Amplitude, (G) Proportion Smooth, (H) Contraction τ, (I) Dilation τ, (J) Pupil Diameter, (K) Main Sequence Slope, and (L) Direction Anisotropy. These plots provide visual evidence of the systematic ordinal differences between binocular vs. monocular viewing.

Figure 3

Plots of oculometric values under non-dominant vs. dominant viewing for (A) initial acceleration and (B) direction noise.

Figure 4

Plots of oculometric values under binocular vs. dominant eye viewing for latency (A), direction asymmetry (B), speed responsiveness (C), saccade rate (D), saccade amplitude (E), proportion smooth (F), pupillary contraction time constant (G), dilation time constant (H), mean pupil diameter (I), and main-sequence slope (J).

Figure 5

Histograms of oculometric values for binocular (B-black), dominant (D-gray), and non-dominant (N-white) eye viewing for (A) initial acceleration, (B) direction noise, (C) latency, (D), saccade rate, (E) saccade amplitude, (F) proportion smooth, (G) main-sequence slope, (H) pupil diameter, (I) dilation τ, (J) contraction τ, and (K) direction asymmetry. Statistical measures are from the statistical analysis above with NS, *, **, ***, **** indicating >0.05, <0.05, <0.01, <0.005, and <0.001, respectively.

Figure 6

Polar plots of mean oculometric values for left and right eye viewing in world/head coordinates (based on the direction of ramp motion in the world) for (A) latency, (B) initial acceleration, (C) steady-state gain, (D) proportion smooth, (E) direction noise, (F) speed noise, and (G) speed responsiveness. 0 deg represents rightward ramp motion (following leftward steps). Error bars represent SEM.

Figure 7

Polar plots of mean oculometric values for dominant and non-dominant viewing in retinal coordinates for (A) latency, (B) initial acceleration, (C) steady-state gain, (D) proportion smooth, (E) direction noise, (F) speed noise, and (G) speed responsiveness. Polar direction represents retinal locus of the stimulus motion during the analysis interval (see legend bottom right). Error bars represent SEM.

Figure 8

Polar plots of mean oculometric values for binocular and monocular viewing with respect to ramp direction in world/head coordinates for (A) latency, (B) initial acceleration, (C) steady-state gain, (D) proportion smooth, (E) direction noise, (F) speed noise, and (G) speed responsiveness. 0 deg represents rightward ramp motion (following leftward directed steps). Error bars represent SEM.