Smooth pursuit

There are two basic methods for recording smooth pursuit eye movements, and eye movement in general. The first is with a search coil. This technique is most common in primate research, and is extremely accurate. An eye movement shifts the orientation of the coil to induce an electric current, which is translated into horizontal and vertical eye position. The second technique is an eye tracker. This device, while somewhat more noisy, is non-invasive and is often used in human psychophysics and recently also in instructional psychology. It relies on the infrared illumination of the pupil to track eye position with a camera.[2]

During oculomotor experiments, it is often important to ensure that no saccades occurred when the subject was supposed to be smoothly pursuing a target. Such eye movements are called catch-up saccades and are more common when pursuing at high speeds. Researchers are able to discard portions of eye movement recordings that contain saccades, in order to analyze the two components separately. Saccadic eye movements differ from the smooth pursuit component by their very high initial acceleration and deceleration, and peak velocity.[3]

The neural circuitry underlying smooth pursuit is an object of debate. The first step towards the initiation of pursuit is to see a moving target. Signals from the retina ascend through the lateral geniculate nucleus and activate neurons in primary visual cortex. Primary visual cortex sends the information about the target to the middle temporal visual cortex, which responds very selectively to directions of movement. The processing of motion in this area is necessary for smooth pursuit responses.[4] This sensory area provides the motion signal, which may or may not be smoothly pursued. A region of cortex in the frontal lobe, known as the frontal pursuit area, responds to particular vectors of pursuit, and can be electrically stimulated to induce pursuit movements.[5] Recent evidence suggests that the superior colliculus also responds during smooth pursuit eye movement.[6] These two areas are likely involved in providing the "go"-signal to initiate pursuit, as well as selecting which target to track. The "go"-signal from the cortex and the superior colliculus is relayed to several pontine nuclei, including the dorsolateral pontine nuclei and the nucleus reticularis tegmenti pontis.[7] The neurons of the pons are tuned to eye velocity and are directionally selective, and can be stimulated to change the velocity of pursuit. The pontine nuclei project to the cerebellum, specifically the vermis and the paraflocculus. These neurons code for the target velocity and are responsible for the particular velocity profile of pursuit. The cerebellum, especially the vestibulo-cerebellum, is also involved in the online correction of velocity during pursuit.[8] The cerebellum then projects to optic motoneurons, which control the eye muscles and cause the eye to move.

Pursuit eye movement can be divided into two stages: open-loop pursuit and closed-loop pursuit. Open-loop pursuit is the visual system's first response to a moving object we want to track and typically lasts ~100 ms. Therefore, this stage is ballistic: Visual signals have not yet had time to correct the ongoing pursuit velocity or direction.[9] The second stage of pursuit, closed-loop pursuit, lasts until the pursuit movement has ceased. This stage is characterized by the online correction of pursuit velocity to compensate for retinal slip. In other words, the pursuit system tries to null retinal velocity of the object of interest. This is achieved at the end of the open-loop phase. In the closed-loop phase, the eye angular velocity and target angular velocity are nearly equal.

Various lines of research suggests a tight coupling for closed loop pursuit and spatial attention. For instance, during the close loop phase selective attention is coupled to the pursuit target such that untracked targets which move in the same direction with the target are poorly processed by the visual system.[10] Recently, a loose coupling of open loop pursuit and attention was suggested, when there is only one possible moving target.[11] This difference between pursuit and saccades may be accounted for by the differences in latency. Pursuit eye movements are initiated within 90-150 ms, while typical latencies for voluntary saccades are in the order of 200-250 ms [12]

Performing smooth pursuit without a moving visual stimulus is difficult,[13] and typically results in a series of saccades. However, pursuit without a visible target is possible under some particular conditions, that show the importance of high-level functions in smooth pursuit maintenance.

If you know which way a target will move, or know the target trajectory (because it is periodic for instance), you can initiate pursuit before the target motion actually starts, especially if you know exactly when the motion will start.[12][14] It is also possible to maintain pursuit if a target momentarily disappears, especially if the target appears to be occluded by a larger object.[14]

Under conditions in which there is no visual stimulation (in total darkness), we can still perform smooth pursuit eye movements with the help of a proprioceptive motion signal (e.g. your moving finger).[15]

Although we can clearly separate smooth pursuit from the vestibulo-ocular reflex, we can not always draw a clear separation between smooth pursuit and other tracking eye movements like the slow phase of the optokinetic nystagmus and the ocular following response (OFR), discovered in 1986 by Miles, Kawano, and Optican,[17] which is a transient ocular tracking response to full-field motion. The latter are both slow eye movements in response to extended targets, with the purpose of stabilizing the image. Therefore, some processing stages are shared with the smooth pursuit system.[18] Those different kinds of eye movements may not be simply differentiated by the stimulus that is appropriate to generate them, as smooth pursuit eye movements can be generated to track extended targets as well. The main difference may lie in the voluntary nature of pursuit eye movements.[19]

Smooth pursuit requires the coordination of many brain regions that are far away from each other. This makes it particularly susceptible to impairment from a variety of disorders and conditions.

• Eye movement
• Eye tracking
• Frontal eye fields
• Saccade
• Superior colliculus
• Endophenotype

• Thier, Peter; Ilg, Uwe J (2005). "The neural basis of smooth-pursuit eye movements". Current Opinion in Neurobiology. 15 (6): 645–52. doi:10.1016/j.conb.2005.10.013. PMID 16271460. S2CID 2261910.

• Pursuit,+Smooth at the US National Library of Medicine Medical Subject Headings (MeSH)