In order to stabilize a moving target’s retinal image, the brain must make continuous visual estimates of target motion and evaluate the trade-off between smoothly modulating eye movement and initiating a saccade. Smooth pursuit eye movement is used for continuous acquisition of visual information, whereas saccadic movement can reduce a large retinal error in a short period of time. Lefèvre and colleagues (2002) introduced the decision rule between pursuit and saccades (Eye-Crossing Time) during one dimensional visual tracking of moving stimuli in humans. Our goal for this study is to expand this notion to investigate if there exists general oculomotor computation for making eye movement decision. In order to achive this goal, three different experimental paradigms have been done in human and monkeys: a 1D and 2D visual tracking with double step-ramp and a single-player version of the video game, Pong. Interestingly, we observed that in the highly predictive situation, such as during the pong game, saccadic eye movement is not captured with the same rule. Here, we apply information theoretic analysis to quantify the interaction between target, gaze, and time.

Are sensory estimates formed centrally in the brain and then shared between perceptual and motor pathways, or is centrally represented sensory activity decoded independently to drive awareness and action? Questions of the brain's information flow pose a challenge because systems-level estimates of environmental signals are only accessible indirectly as behavior. Assessing whether sensory estimates are shared between perceptual and motor circuits requires comparing perceptual reports with motor behavior arising from the same sensory activity. Extrastriate visual cortex mediates both the perception of visual motion and it provides the visual inputs for behaviors like smooth pursuit eye movements. Pursuit has been a valuable testing ground for theories of sensory information processing because the neural circuits and physiological response properties of motion-responsive cortical areas are well-studied, sensory estimates of visual motion signals are formed quickly, and the initiation of pursuit is closely coupled to sensory estimates of target motion. Here we analyze variability in visually-driven smooth pursuit and perceptual reports of target direction and speed in human subjects while we manipulate the signal to noise level of motion estimates. Comparable levels of variability throughout viewing time and across conditions provide evidence for shared noise sources in the perception and action pathways arising from a common sensory estimate. We find that conditions that create poor, low-gain pursuit create a discrepancy between the precision of perception and that of pursuit. Differences in pursuit gain arising from differences in optic flow strength in the stimulus reconcile much of the controversy on this topic.

In pursuit, visual estimates of retinal image motion are translated to motor commands to smoothly counter-rotate the eye in order to stabilize the retinal image of a moving target. The visual inputs for pursuit behavior arise in the middle temporal cortical area (MT) where neuronal responses are tuned for motion direction and speed. Motion direction discrimination in pursuit is approximately a factor of ten lower than MT neuron discrimination thresholds, suggesting that visual motion estimates are derived from cortical population responses. In previous work, we have shown that variability in the initiation of pursuit arises from variability in motion estimation, as if visual estimates are decoded from cortical inputs with errors, but are then loyally translated to movement by the oculomotor system. Here we show that while a variety of population decoding models can reproduce the accuracy of pursuit eye speed and direction, reproducing the trial-to-trial eye movement variation places strong constraints on the coordinate frame and number of free parameters in the decoding model. By extracting the principal components of the deviations from trial-averaged mean eye velocity for each trial of pursuit behavior, we estimate the trial error in speed, direction, and movement onset time. We find that direction and speed errors in pursuit were significantly more highly correlated at oblique directions than at cardinal directions, a pattern replicated by adding noise in motor coordinates aligned with the pulling direction of the eye muscles (i.e. horizontal, vertical). Initial results indicate that there is no such anisotropy in direction-speed error correlations in perception, as would be expected with an effect due to added motor noise. However, using both correlated and uncorrelated Poisson model neurons coupled to experimental measurements of actual pursuit behavior, we show that the most parsimonious decoding models operate in visual coordinates (direction, speed) rather than in motor coordinates.

Performance in sensory-motor behaviors guides our understanding of many of the key computational functions of the brain: the representation of sensory information, the translation of sensory signals to commands for movement, and the production of behavior. Eye movement behaviors have become a valuable testing ground for theories of neural computation because the neural circuitry has been well characterized and eye movements can be tightly coupled to cortical activity (Osborne et al., 2005-9). Here we show that smooth pursuit eye movements, and the cortical sensory signals that mediate them, demonstrate the hallmarks of efficient sensory coding. Barlow (1961) proposed that neurons should adapt their sensitivity as stimulus conditions change in order to maintain efficient coding of sensory inputs. Evidence for efficient coding of temporal fluctuations in visual contrast has been observed in the retina, lateral geniculate nucleus, and visual cortex (Wark et al., 2009, Sharpee et al., 2006). We asked whether adaptation to stimulus variance generalizes to higher cortical areas whose neurons respond to features of visual signals that do not drive adaptation in the sensory periphery, and whether such adaptation impacts performance of visually-driven behavior. Specifically, we have studied the impact of dynamic fluctuations in motion direction on the gain of smooth pursuit and found neural correlates of pursuit adaptation in cortical area MT.

In order to make appropriate gaze decisions, the brain must integrate information across the visual field to identify and locate objects, estimate motion, and allocate attention. We simplify and decompose this complex problem by asking how visual motion is integrated across the visual field to drive smooth pursuit eye movements. The pursuit system has become a valuable testing ground for theories of sensory estimation and motor control because the neural pathways are well described and the relationship between motion estimation and pursuit initiation is very precise. Eye direction and speed can be a loyal representation of internal estimates of target motion in the absence of experienced-based modulations of the oculomotor state. In past work, we have demonstrated that a linear model of motion integration across time accounts for a high percentage of the variance in eye velocity and provides a good description of temporal filtering in the visual motion inputs for pursuit. Here we expand upon that research to study both spatial and temporal motion integration visual stimuli, asking how motion signals are weighted across the visual scene to drive smooth eye movements. We have analyzed pursuit of visual stimuli with spatially and temporally distributed motion directions and speeds in humans. We find that a spatially and temporally restricted linear filter captures the relationship between motion signals and pursuit eye movements, accounting for 58% of the variation in eye direction during pursuit initiation. The variation not captured by the best linear model is Gaussian distributed, indicative of noise in visual-motor processing. We find that motion within ~6° of the fixation point is integrated over ~100 ms to drive pursuit in humans. The linear filter generalizes across stimulus forms that differ in their amount of motion energy. These results suggest that relatively simple estimation rules may underlie the processing of moving scenes in natural gaze behavior.