Yasushi Kodaka

The role of MST neurons during ocular tracking in 3D space.

Whenever we move around in the environment, the visual system is confronted with characteristic patterns of visual motion, termed optic flow. The optic flow contains important information about self-motion and the 3D structure of the environment as discussed in other chapters in this book. O n the other hand, visual acuity is severely affected if the images of interest on the retina move at more than a few degrees per second. A major function of eye movements is to avoid such retinal motion and thereby improve vision. The observer’s movements activate the vestibular organs and are then compensated by vestibuloocular reflexes. However, the vestibuloocular reflexes are not always perfect, and the residual disturbances of gaze are compensated by the visual tracking system(s). Until recently, visual stabilization of the eyes was regarded only in terms of providing’backup to the canal-ocular vestibular reflexes, which deal solely with rotational disturbances of the observer. This is reflected in the stimulus traditionally used to study these visual mechanisms: The subject is seated inside a cylinder with vertically striped walls that are rotated around the subject at constant speed, often for periods of a minute or more. Because of the cylinder’s bulk, it is usual to first bring the cylinder up to speed in darkness and to then suddenly expose the subject to the motion of the stripes by turning on a light. The ocu-

Short-latency vergence eye movements elicited by looming step in monkeys.

The looming associated with forward and backward motion of the observer has been shown to elicit vergence eye movements with short-latency (approximately 80 ms) in human subjects. We studied the vergence eye movements elicited by looming in three monkeys (Macaca fuscata). The animals faced a large tangent screen onto which a random dot pattern was back-projected. The movements of both eyes were recorded with an electromagnetic induction technique. Fifty milliseconds after a centering saccade, this first pattern was replaced with a new one that showed the same image viewed from a slightly different distance. This looming step (two-frame movie) included both radial optic flow and a size change. As expected from the optical geometry, centrifugal flow coupled with enlargement (signaling forward motion) increased the vergence angle, whereas the converse combination decreased the vergence angle. In both cases, the optimal step-change in apparent viewing distance was 2%. The latency of these vergence responses was very short and similar to those induced when disparity steps are applied to such large patterns (approximately 60 ms). We suggest that these two systems act in synergy to help maintain binocular alignment during forward and backward motion of the observer.

Preparatory modulation of the gain of visuo-motor transmission for smooth pursuit in monkeys

Brief movement of a foveated target is known to elicit higher velocity ocular (tracking) responses if the target is in motion rather than stationary. We determined whether similar perturbations of a stationary target would have greater ocular effects if we merely increased the probability that the target might undergo sustained motion. For this, we examined the effect of interleaving trials in which the target was always stationary with trials in which the target underwent sustained motion that required the animal to track. We found that perturbation of the stationary target had a greater effect when there were interleaved trials in which the target moved, as though the gain of the visuo-motor transmission had been increased in anticipation of future tracking.

Effect of target saliency on human smooth pursuit initiation: interocular transfer

Our previous study showed that the saliency of a target increases the gain of smooth pursuit initiation. In this study, we examined the interocular transfer of this effect in five humans. A square red frame surrounding the target was used as a cue to indicate the initial target position. In the cue condition, the responses were similar, irrespective of the eye to which the cue was presented, and were significantly larger than in the no-cue condition. The result suggests that central pathways that receive input from both eyes mediate the effect of saliency on smooth pursuit initiation.

Vergence responses to forward motion in monkeys: visual modulation at ultra-short latencies.

In two monkeys, we measured the initial vergence eye movements elicited by sudden forward motion on a linear sled. Animals faced a tangent screen and experienced the translation while in darkness, fixating a small, centered spot, or viewing a large-field pattern. Forward movements elicited convergent responses that were enhanced in the presence of the visual stimuli after a latent period. The enhancement was greater with the large-field pattern than with the small spot. The latencies of these visual effects were ultra-short and less than those reported for the vergence eye movements elicited by pure visual stimuli when applied in isolation. It is possible that these ultra-short latencies resulted from the fact that there were multiple visual cues available to sense the change in viewing distance, including binocular disparity, radial optic flow, size changes, and blur. Another possibility is that the very earliest visual effects during forward motion resulted from direct modulation of the otolith-mediated, translational vestibulo-ocular reflex.

Vertical ocular responses to constant linear acceleration generated by fore-aft head translation in monkeys.

We examined the vertical linear vestibuloocular reflexes (LVORs) elicited by constant linear acceleration (0-0.5 g for >95 ms) during transient fore-aft translation in three monkeys. In the dark condition, small but consistent downward ocular responses to forward translation were observed (latencies >41 ms) when the initial vertical eye positions were at 0 degrees , although eye movements following backward translation were inconsistent among animals. These downward ocular responses showed the following three characteristics: they were independent of vertical gaze eccentricities, their magnitudes were almost proportional to the forward acceleration, and they were reduced by the large-field (not the spot) visual information. These characteristics revealed that the downward ocular responses to forward translation were the tilt LVORs. In addition, we recognized that the translational LVOR, which depended on vertical gaze eccentricities, was working at the same time. Our data suggest that constant linear acceleration during forward translation evokes the tilt LVOR, as well as the illusory tilting perception.

Effects of a Large‐Field Visual Scene on the Vergence Response to Naso‐Occipital Linear Motion in Monkeys

Vergence eye movements are elicited by visual stimuli1,2 as well as naso-occipital (NO) linear head motion.3 Therefore, vergence eye movements during NO linear motion can be influenced by visual stimuli. As the first step in understanding the role of visual stimuli in the vestibular vergence response, we studied the effects of a large-field visual scene on the oculomotor response to forward linear motion in monkeys.

Whole-Body Roll Tilt Influences Goal-Directed Upper Limb Movements through the Perceptual Tilt of Egocentric Reference Frame

In our day-to-day life, we can accurately reach for an object in our gravitational environment without any effort. This can be achieved even when the body is tilted relative to gravity. This is accomplished by the central nervous system (CNS) compensation for gravitational forces and torque acting on the upper limbs, based on the magnitude of body tilt. The present study investigated how performance of upper limb movements was influenced by the alteration of body orientation relative to gravity. We observed the spatial trajectory of the index finger while the upper limb reached for a memorized target with the body tilted in roll plane. Results showed that the terminal location of the fingertip shifted toward the direction of body tilt away from the actual target location. The subsequent experiment examined if the perceived direction of the body longitudinal axis shifted relative to the true direction in roll plane. The results showed that the perceived direction of the body longitudinal axis shifted toward the direction of the body tilt, which correlated with the shift of the terminal location in the first experiment. These results suggest that the dissociation between the egocentric and gravitational coordinates induced by whole-body tilt leads to systematic shifts of the egocentric reference frame for action, which in turn influences the motor performance of goal-directed upper limb movements.

Neuronal Activity in Monkey Cortical Area 6 during the Initial Phase of Smooth Pursuit against a Stationary Background

In many of the studies on pursuit eye movement, subjects tracked a moving spot in the dark. In the natural world, however, we usually track targets against a textured background. Laboratory experiments indicate that primates can pursue a target against a textured background, although some reduction in pursuit gain and/or increase in latency has been reported (for review, see Ilg1). Somewhere in the brain, there must be a system that can select the target from the background, perhaps involving a form of selective attention. In an attempt to understand some of the mechanisms involved, we have begun to study the effects of a stationary background on neuronal activity in area 6 of monkey frontal cortex during the initial phase of smooth pursuit.