R EVIEW How are visual areas of the brain connected to motor areas for the sensory guidance of movement? Mitchell Glickstein Visual areas of the brain must be connected to motor areas for the sensory guidance of movement.The first step in the pathway from the primary visual cortex is by way of the dorsal stream of visual areas in the parietal lobe.The fact that monkeys can still guide their limbs visually after cortico–cortical fibres have been severed suggests that there are subcortical routes that link visual and motor areas of the brain.The pathway that runs from the pons and cerebellum is the largest of these. Pontine cells that receive inputs from visual cortical areas or the superior colliculus respond vigorously to appropriate visual stimuli and project widely on the cerebellar cortex. A challenge for future research is to elucidate the role of these cerebellar target areas in visuo–motor control. Trends Neurosci. (2000) 23, 613–617 OST HUMAN and animal movements are Munk4 discovered that unilateral lesions of the stri- Munder continual sensory guidance. In humans ate cortex of the occipital lobe in monkeys produced and the higher primates, the dominant sense that hemianopia, that is, the monkey became blind in controls movement is vision. One of the central the visual field on the opposite side to the lesion. problems in neuroscience is to understand how the Munk’s identification of the visual functions of the brain uses visual input to control whole body move- occipital lobe was confirmed in humans by ments, such as walking through a crowded railroad Henschen5. Henschen identified the cortex on the station without bumping into other people or run- banks, lips and depths of the calcarine fissure as the ning to catch a ball. What are the pathways in the crucial locus of lesions that are associated with brain that are involved in more delicate movements, hemianopia, because damage to cortex outside this such as threading a needle or picking ripe fruit from area failed to produce hemianopia. However, the a vine? When analyzed in detail, what might appear lesions in Henschen’s patients were large and dif- to be a simple directed action might, in fact, consist fuse, which made it difficult for him to establish pre- of several independent components. Jeannerod and cisely how visual fields are represented on the cor- colleagues1, for example, showed that there are at tex. Henschen was correct in identifying the left least two independent mechanisms that control occipital lobe as receiving its input from the left half what appears to be a simple, integrated act of reach- of each retina, hence the right half of the visual ing out to grasp an object. Sakata and colleagues2 field. He also correctly identified the upper bank of have identified an area, the anterior interparietal the calcarine fissures as receiving its input from the area (AIP), within the rostral bank of the intrapari- upper retina, hence the lower visual field. However, etal sulcus, which contains neurones that extract Henschen believed (incorrectly) that the central detailed visual information about the nature of the visual fields were represented anteriorly within the object to be grasped and that is reciprocally con- calcarine fissure. Smaller lesions were required to nected to a motor area within the frontal lobe. determine the precise way in which the visual fields Although the location of the primary visual areas are mapped. Inouye6,7 studied the visual deficits in and motor areas of the cerebral cortex has been brain injured soldiers of the Russo–Japanese war. In known for more than a century and a great deal is many of his patients, a small bullet penetrated the now known about the further processing of the skull at high velocity, leaving a straight track of visual input and the interconnections and projec- injury through the brain. Inouye could therefore use tions of the motor areas3, little is known about how the location of the entrance and exit wounds to cal- these areas are linked. Ultimately, cortical visual culate where the bullet would have crossed the areas must be connected to descending pathways for visual cortex or interrupted the visual radiations. He the control of movement. Although there are inter- studied the visual field defects in these soldiers and Mitchell Glickstein connections between the parietal lobe visual and accordingly constructed a map of how the visual is at the Dept of frontal motor areas, visual and motor cortical areas fields are represented on the striate cortex. The cen- Anatomy, are also strongly linked by way of a pathway tral fields were represented caudally and occupied a University College through the cerebellum. larger cortical area than the peripheral fields. London, Gower The primary sensory and motor areas of the ce- Holmes and Lister8, studied British casualties in Street, London, UK rebral cortex were first identified in the 19th century. the First World War and provided a more detailed WC1E 6BT. 0166-2236/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S0166-2236(00)01681-7 TINS Vol. 23, No. 12, 2000 613 R EVIEW M. Glickstein – Sensory guidance of movement representation of the mapping of the visual fields on Although the current terminology is more sophis- the human striate cortex, which has been repeatedly ticated, the assumption that a series of cortico–cor- confirmed since that time. tical fibres links sensory areas to motor areas of the Identification of the motor area of the cerebral cortex still prevails. There are no direct anatomical cortex also began with animal experiments. Fritsch projections from primary visual cortex (Area 17) to and Hitzig9 showed that weak electrical stimulation the motor and premotor areas (Areas 4 and 6). of a specific region of the dog’s frontal cerebral cor- Instead, cortico–cortical links between the primary tex could elicit movement of the face or limb on the visual and motor areas are indirect and the pathway contralateral side of the body. If electrical stimula- is typically thought to involve a series of tion of a given point elicited limb movement, abla- cortico–cortical links that originate in Area 17 and tion of the cortex at that point hindered the use of end in Areas 4, 6 or 8 for the visual guidance of the that limb. Shortly thereafter, Betz10 described the fingers, limbs or eyes. The corpus callosum is characteristic large pyramidal cells in layer V of cor- thought to mediate tasks in which the visual cortex tex that are associated with the motor area. The on one side had to be linked to the motor cortex on interpretation of these experiments as indicating a the opposite hemisphere. motor-control area in the cerebral cortex was ini- The dorsal stream of extrastriate visual areas tially controversial because some authors insisted that the stimulation must have caused a sensation, In 1982, Ungerleider and Mishkin proposed12 that and it was that sensation which led to movement. the extrastriate visual areas could be broadly segre- Nevertheless, by 1900 the identity and location of gated into two streams on the basis of anatomical the primary sensory and motor cortical areas were and behavioural evidence: one dorsal stream, cod- widely accepted. ing the location of visual objects (‘where is it?’) and Although the evidence for localization of visual and the other ventral stream, identifying the nature of motor functions was clear, surprisingly few questions the object (‘what is it?). In the same year, Jack May were asked regarding how sensory areas might be and I13 proposed that a similar dichotomy exists linked to motor areas for the sensory guidance of between dorsal and ventral extrastriate cortical movement. Most authors assumed that there must be visual areas on the basis of the differential projec- a series of cortico–cortical connections that arise in tions of these two areas to the pontine nuclei. The the visual cortex and end in the motor cortex. Figure 1 dorsal, parietal areas project heavily to the cerebel- reflects this generally accepted view. lum by way of the pontine nuclei, whereas the ven- tral stream has little or no such projection. We sug- gested that an important function of the parietal lobe visual areas might be in the visual control of movement, based on its anatomical connections to the cerebellum (Fig. 2). Although visually guided movement of the wrist R a c R and fingers of monkeys is impaired following par- ietal lobe lesions, control lesions of the temporal lobe or peri-arcuate cortex produce little or no a impairment. The evidence is now clear that the first a step in the pathway that links visual to motor corti- c cal areas is by way of the dorsal stream of extrastri- ate visual areas, because lesions of these areas in fe fe monkeys profoundly impair visual guidance of movement14. By contrast, lesions of cortical areas in fa fa the ventral stream can impair form, colour or face recognition, but do not affect visuo–motor coordi- nation15. Indeed lesions of the dorsal stream visual areas are so severe in their effects on visual control of movement that Ferrier16,17 initially misinterpreted these effects in monkeys as blindness (Fig. 3). Almost identical deficits in visually guided movement are present in humans after lesions of analogous extra- striate cortical visual areas18,19. Connections of the dorsal stream visual areas k How do the dorsal stream visual areas connect to cortical and subcortical brain areas that control movement? Although there are indirect cortico– trends in Neurosciences cortical routes from these areas to Areas 4 and 6, monkeys can still execute rapid and accurate visually Fig.