Smooth ocular pursuit stabilizes the image of an object on or near the fovea (see glossary) for optimal visual acuity during slow movement of the object or of the body. Smooth pursuit is needed to hold the eye on a stationary target during locomotion. When a target located off to one side is viewed during locomotion, smooth pursuit holds its image at the fovea, despite relative motion of the background (Miles, 1993). Retinal image motion exceeding 5–6 deg/s reduces visual acuity for the higher spatial frequencies of high contrast objects (Burr and Ross, 1982, Murphy, 1978). Defective generation of smooth tracking has been a well-recognized clinical sign of cerebral lesions since early in the last century (Fox & Holmes, 1926). Information from anatomical and physiological experiments on monkeys and non-primate animals serves as a contemporary foundation for understanding ocular motor systems. Quantitative recordings of smooth pursuit in normal human subjects, and in patients with focal brain lesions identified by high resolution brain imaging, together with functional imaging of the brain during pursuit of a foveal target have advanced our knowledge considerably. Functional imaging reveals areas that are active during eye motion, whereas correlation of impaired function with lesion sites can identify the brain regions that are critical for the motion. The neurophysiology and anatomy of smooth pursuit in animals and brain imaging during pursuit in humans are discussed in other articles of this issue. Here the effects of lesions of the brain on smooth pursuit are discussed.
The cardinal clinical sign of impaired smooth pursuit is the presence of saccadic, rather than smooth, pursuit when a slowly moving object is tracked. When smooth eye movement velocity fails to match the object’s speed, catch-up saccades are dispatched to place the foveas of both eyes on the target (Fig. 1). The saccadic movements make the pursuit appear irregular, giving rise to “cogwheel” pursuit. In the clinic a target, the examiner’s finger or pen tip, moving at a speed of about 20 deg/s suffices to detect saccadic, rather than smooth, tracking. Targets can usually be followed smoothly (with eye velocity approximating target velocity) with few saccades provided that the target velocity is less than about 50 deg/s, and, for periodic targets moving in sine waves, the frequency of oscillation is less than about 1 Hz. Some human subjects can attain smooth pursuit eye speeds of up to 150 deg/s for sinusoidal target motion and 100 deg/s for constant velocity motion.Detection of defective smooth pursuit
- •
The cardinal clinical sign of impaired, or paretic, smooth pursuit is saccadic tracking of a slowly moving object.
- •
Saccadic pursuit consists of catch-up saccades that foveate the target and compensate for low velocity smooth eye movements.
- •
Smooth pursuit is best measured by its gain, the ratio of smooth eye movement velocity to target velocity.
- •
Saccadic intrusions such as square wave jerks and anticipatory saccades should not be mistaken for the compensatory catch-up saccades that are interspersed among low velocity smooth eye movements.
Although saccadic pursuit can be detected by clinical examination, measurement of defective smooth pursuit requires quantitative oculographic study of smooth pursuit gain, the ratio of smooth eye movement velocity to target velocity (Box 1 and glossary). The gain required for optimal vision is near unity. Catch-up saccades may be counted as an alternate indicator of low smooth eye movement speed, but saccadic intrusions, such as anticipatory saccades (see glossary) and square wave jerks (see glossary) (Sharpe & Fletcher, 1984), can be mistaken for actual catch-up saccades (see glossary) in the direction of tracking. Saccadic intrusions can confound the detection of genuine lowering of smooth pursuit speed for which catch-up saccades compensate by briefly foveating the moving target. Square wave jerks are intrusions of paired horizontal saccades; each intrusion consists of a small amplitude saccade that take the fovea off target followed after an interval of about 200 ms by a saccade that returns to the target. Frequent square wave intrusions typically signify diffuse of cerebral hemispheric or cerebellar (see glossary) or basal ganglia (see glossary) disease (Sharpe, Herishanu, & White, 1982). Anticipatory saccades are intrusions that move the eyes off target toward another target or the expected future location of the viewed target (Fletcher & Sharpe, 1986) (Fig. 1 illustrates an anticipatory saccade among catch-up saccades). Saccadic intrusions occur during attempted steady fixation of a stationary target, as well as during smooth tracking of a moving target. In contrast to catch-up saccades, they are not a manifestation of smooth pursuit dysfunction per se. Although the detection of saccadic intrusions is can indicate diffuse or focal brain disease during both attempted steady fixation and pursuit (Sharpe et al., 1982), they must be distinguished from genuine catch-up saccades which compensate for lowered smooth eye movement speeds. Thus determination of actual smooth eye movement velocity, rather than the frequency of saccades, provides optimal measurement of impaired smooth pursuit.
The pathological nature of lesions is important in appraising their effects. After several days Infarcts and surgical resections typically have destroyed tissue within the imaged lesion, and a deficit can be attributed to the lesion. Furthermore normal function after the lesion provides evidence that the imaged lesion region is not essential for the function. However, in the case of hemorrhages, neoplasms, and vascular malformations displacement, edema, and distortion of tissue can leave function intact within the lesion boundary. After such lesions no anatomical conclusion can be drawn from a normal function, although loss or impairment of a measured function permits conclusion that the region is critical for the function. The technique of determining the overlap of imaged lesions from a group of patients who have defective smooth pursuit provides evidence that the lesion overlap region shared by all patients is critical for the pursuit function. Conversely, normal pursuit after infarcts or surgical resections provides evidence that the damaged region is not crucial.
The acuteness of lesions is imperative in assessing roles of brain regions. Low gain pursuit, as measured in the laboratory or saccadic tracking as seen in the clinic or at the bedside, may be apparent only in the first few hours or days after a lesion occurs. Recovery may signify healing the pathological process, or compensatory function of adjacent or remote brain regions, or redundancy of pursuit functions in different pathways. The later may indicate that the damaged region participates in pursuit but is not crucial. On the other hand persistence of low gain smooth pursuit with chronic focal lesions indicates that the damaged areas are necessary for normal pursuit. Chronic degenerative or metabolic diseases of the brain have diffuse effects and often impair smooth tracking; those conditions damage multiple regions that participate in pursuit and consequently provide less information of localizing value for neurological diagnosis. Nonetheless, low gain smooth pursuit is a sensitive sign of chronic diffuse brain disease that can be useful in the detection of neurological disorders.
Although saccadic tracking in one or more directions is the diagnostic sign of paretic smooth pursuit, damage to different brain regions can produce different patterns of defect. These patterns are presented in this article where the anatomic region responsible for the type of defect is discussed. Notably, retinotopic and craniotopic smooth pursuit defects are identified only with cerebral hemispheric lesions, and contraversive pursuit defects are detected after some caudal brainstem lesions, whereas ipsiversive pursuit defects result from focal lesions at each level of the intracranial neuraxis. This presentation is most practical in the context of anatomical diagnoses.
Regulation of smooth pursuit involves the operations of extensive focal neuronal regions and pathways in the cerebral hemispheres, cerebellum and brainstem. These regions and circuits are discussed below. Briefly, in the cerebral hemispheres, the junction of parietal, temporal and occipital cortex at area V5, and its connections with parietal cortex and its parietal eye field, and the frontal eye fields (FEF) (see glossary), process smooth eye movement commands that are conveyed to the brainstem and cerebellum. Two parallel cerebral cortico-ponto-cerebellar circuits project to the cerebellum: A FEF-to-reticularis tegmenti pontis (NRTP)-to-cerebellum, and a visual area V5-to-dorsolateral pontine nucleus (DLPN) (see glossary)-to-cerebellum pathway govern smooth pursuit. Crucial cerebellar regions are the ventral paraflocculus and lobules VI and VII of the vermis which in turn project to the vestibular nuclei. Direct connections from the vestibular nuclei to ocular motor nuclei complete the principal pursuit circuit in the brainstem. Other anatomical elements of the pursuit system in the cerebral hemispheres, cerebellum and brainstem are also considered in the following sections.