Electrical stimulation of the frontal eye field in a monkey produces combined eye and head movements.

The frontal eye field (FEF), an area in the primate frontal lobe, has long been considered important for the production of eye movements. Past studies have evoked saccade-like movements from the FEF using electrical stimulation in animals that were not allowed to move their heads. Using electrical stimulation in two monkeys that were free to move their heads, we have found that the FEF produces gaze shifts that are composed of both eye and head movements. Repeated stimulation at a site evoked gaze shifts of roughly constant amplitude. However, that gaze shift could be accomplished with varied amounts of head and eye movements, depending on their (head and eye) respective starting positions. This evidence suggests that the FEF controls visually orienting movements using both eye and head rotations rather than just shifting the eyes as previously thought.

Architecture of a gain controller in the pursuit system.

A monkey can pursue faster target oscillations if they appear during ongoing smooth pursuit than if they appear while the monkey is fixating a stationary target. Others have proposed a switch in the pursuit circuit to account for this bistable sensitivity to high frequency targets. It is hypothesized that the switch is closed only during pursuit, permitting the retinal motion signal to pass through the circuit at full gain. Losses in pursuit gain caused by certain cortical lesions do mimic the effect of a switch jammed open. To explore this gain adjustment mechanism further, we measured in monkeys the smooth eye movements in response to a high frequency sinusoidal target (called 'humm') presented under a variety of testing conditions. Pursuit gain measured in response to this humm was not merely bistable. Rather, a graded gain modulation of the pursuit system was possible. Furthermore, the gain adjustment had some directional sensitivity to it, enhancing the response to humm along one axis more than the other. In exploring the factors which gated the gain adjustment, it appeared that the movement of the eyes and not the image motion that occurs during pursuit was paramount for enhancing pursuit gain. Gain was not enhanced by saccadic but only by smooth pursuit tracking movements. Finally, gain could be modulated somewhat by covert signals such as the expectation of future smooth pursuit movements.

Ablation of the pursuit area in the frontal cortex of the primate degrades foveal but not optokinetic smooth eye movements.

1. Neural pathology which impairs foveal smooth pursuit eye movements typically also degrades optokinetic pursuit of large textures, suggesting that the two kinds of pursuit share a common circuit. This study reports an exception. After sequential bilateral ablation of the pursuit area in the frontal lobe three monkeys displayed degraded pursuit of a small foveal target but performed normally on identical measures of optokinetic pursuit. 2. A related experiment in one subject demonstrated a pursuit deficit when the foveal target moved relative to the background, but not when background and target moved together. The frontal pursuit area may specifically control pursuit of relative motion, and do so by receiving signals primarily from motion detectors in the macular part of the visual field.

Monkeys and mug shots: cues used by rhesus monkeys (Macaca mulatta) to recognize a human face.

We investigated the cues rhesus monkeys (Macaca mulatta) use to recognize a familiar human face. To manipulate facial cues, schematic faces were constructed with Identi-Kit materials derived from mug shots. The monkeys (N = 4) spontaneously classed Identi-Kit as faces on initial presentations. The monkeys then learned to distinguish one Identi-Kit face, the standard, from others. Panel presses indicated recognition of the standard face. Eye movement recordings revealed that the monkeys predominantly fixated on the eyes of the standard face. When the standard face was transformed by removing, altering, or reorienting its features, only alterations of eyes or brows lowered recognition; removal of eyes, brows, nose, or lips did not. Responses to rotated, inverted, and scrambled versions of the standard face varied but generally disrupted recognition. We concluded that features and configuration were used to recognize the human face.

Lesions of the frontal eye field impair pursuit eye movements, but preserve the predictions driving them.

Visually guided eye movements are driven by a mix of current signals (e.g. visual motion) and prior experience (predictive strategies). Previously, large ablations of the frontal eye field (FEF) impaired visually guided smooth pursuit. This study examined if the pursuit decrement could be laid to a selective loss in predictive signals. Normal monkeys demonstrated some of the same predictive eye movements documented in humans. They pursued periodic visual targets with near-zero phase lag. When such targets suddenly disappeared, the monkeys continued smooth pursuit without visual guidance for several reaction times. These epochs of 'blind pursuit' achieved a peak velocity proportional to the prior target frequency just experienced by the monkey. With predictable step-ramp targets smooth eye movements sometimes preceded target motion (anticipatory pursuit). Reaction time to begin pursuit was influenced by the target velocity of prior trials. Small unilateral ablations of 'low threshold' FEF showed smooth pursuit if the fundus of the arcuate sulcus was thoroughly removed. On the step-ramp targets the slowing was evident in both the initial 100 ms and subsequent portions of pursuit. During sine pursuit blind epochs were more slowed by surgery than were visually guided epochs of pursuit. Ipsilateral anticipatory pursuit was abolished in one subject, but not in the other subject with pursuit deficits. Otherwise, pursuit after surgery continued to display the influence of predictive strategies. The average phase lag of periodic pursuit remained much less than the pursuit system's reaction time. Blind epochs persisted after the periodic target disappeared, albeit with a lower peak velocity. Reaction time to begin pursuit of step-ramp targets remained a function of the monkey's experience on prior trials. It is argued that the FEF pursuit deficits do not reflect the loss of visual motion signals or the loss of some 'cognitive' signal such as a prediction about the target's motion, but rather can be explained as a motor deficit.

Frontal eye field lesions impair predictive and visually-guided pursuit eye movements.

The study initially explored the frontal eye field's (FEF) control of predictive eye movements, i.e., eye movements driven by previous rather than current sensory signals. Five monkeys were trained to pursue horizontal target motion, including sinusoidal targets and "random-walk" targets which sometimes deviated from a sine motion. Some subjects also tracked other target trajectories and optokinetic motion. FEF ablations or cold lesions impaired predictive pursuit, but also degraded visually guided foveal pursuit of all targets. Unilateral lesions impaired pursuit of targets moving in both horizontal orbital fields and in both directions of movement. Saccadic estimates of target motion were generally accurate. The slow-phase velocity of optokinetic pursuit (collected after 54 s of OKN) also appeared normal. Pursuit recovered over 1-3 weeks after surgery but the deficits were then reinstated by removal of FEF in the other hemisphere. Thereafter, a slight deficit persisted for up to 10 weeks of observation in two subjects. The pattern of symptoms suggests that FEF lies subsequent to parietal area MST and prior to the pontine nuclei in controlling pursuit eye movements.

Saccadic disorders caused by cooling the superior colliculus or the frontal eye field, or from combined lesions of both structures.

We used reversible cold lesions to explore the oculomotor consequences of separate and combined dysfunction of the superior colliculus (SC) and the frontal eye field (FEF). Two monkeys were trained to fixate visual targets. In one we measured visually driven saccades while cooling the right SC, first alone, then in combination with bilateral FEF ablation. Two cryodes in the other subject permitted measurement of eye movements during cooling of either the right FEF or the right SC, or both structures together. Cooling FEF mainly caused a neglect. Raising the cryode temperature slightly alleviated the neglect and uncovered a subtle saccadic deficit. It consisted of a slight reduction in saccadic amplitude and increase in saccadic reaction time. Cooling the SC alone lengthened saccadic reaction time and reduced saccadic amplitude more dramatically, causing the monkeys' initial saccade to miss the target. Some correction occurred but a targeting error persisted to the end of the trial. Combined lesions of FEF and SC greatly increased reaction times, reduced saccadic amplitude, and caused large and persistent targeting errors. The changes in saccadic amplitude and the targeting errors were a function of the monkey's eye position. Combined lesions also truncated the ocular range of the monkeys.

Disconnection of parietal and occipital access to the saccadic oculomotor system.

The experiment explored the networks through which signals arising from visual areas of cortex control saccadic eye movements. Electrical stimulation of the inferior parietal and the occipital cortex (here termed the "posterior eye fields") normally evokes saccadic eye movements. We replicated previous reports that these evoked eye movements ceased after large tectal ablations. This initial finding suggested that the "posterior eye fields" depended on a single route of access to the saccade generator, one descending through the superior colliculus (SC). On closer examination, the critical lesion appeared to be one which removed the SC and cut efferents from the frontal eye field (FEF) coursing nearby. Subsequently we confirmed that eye movements evoked from the posterior eye fields ceased after cooling the SC, or cutting its efferents- but only when one of these procedures was combined with FEF ablation. Thus, visual signals from the occipital and inferior parietal cortex have more than one, but perhaps only two routes of access to the oculomotor system. One passes through the superior colliculus, the other through the frontal eye field. Ancillary experiments revealed that inferior parietal and FEF ablations, alone or combined, do not disrupt saccades evoked from the occipital lobe. Striate and prestriate areas can therefore use their own direct input to the SC or to the basal ganglia to drive saccadic eye movements.

Targeting errors and reduced oculomotor range following ablations of the superior colliculus or pretectum/thalamus.

The physiology of the superior colliculus (SC) implicates it in the visual control of eye movements. In the primate, acute inactivation of the superior colliculus delays the onset of a visually guided saccade, slows its velocity, and shortens its amplitude. Previous research leaves uncertain whether other oculomotor disorders which sometimes follow ablation of this structure are due to tectal pathology, to neural damage surrounding the tectum, or to both causes. In this study, 7 cynomolgus monkeys received SC ablations. In 3 others, control lesions were placed in the pretectal/posterior thalamic region. Both procedures produced a qualitatively similar syndrome of 4 oculomotor changes. Reaction time to initiate saccades to visual targets was slowed. Secondly, the surgery constricted the normal ocular range. At the worst, movement was confined to a radius of 10-12 degrees of primary gaze. The monkeys displayed two kinds of inaccuracies when attempting to foveate stationary visual targets within their surviving ocular range. Saccadic amplitude was reduced, causing the monkeys' initial attempt to fall short of foveating the target. If the target remained lighted there then ensued a series of stepwise corrective saccades toward it. The corrective saccades ceased with the eyes still at a position short of the target. Eye position remained in error for the duration of the trial. The final position was independent of the target's retinal position or the vector of the motor command needed to acquire the target. Rather, the error was related to the angular position of the target about the head ( = desired eye position). The syndrome appeared qualitatively similar whether resulting from tectal or the more rostral pretectal/diencephalic ablation. When occurring along the horizontal axis, the deficits appeared to require damage to the superior colliculus, perhaps combined with pathology of some other structure. The same syndrome along the vertical axis was better correlated with pretectal/diencephalic pathology. Invasion of these areas together with invasion of the transthalamic axons from the frontal eye fields is interpreted as the critical pathology responsible for the syndrome. A similar oculomotor trajectory can be modelled by supposing a loss in the gain of the signal which conveys the target's retinal position, combined with one other fault in the circuit: either a loss in gain of the eye position signal, or the signal representing the target's position in craniocentric coordinates.

Alterations of retinal inputs following striate cortex removal in adult monkey.

The morphology of the retina and central retino-recipient nuclei was studied in two monkeys that had undergone total bilateral striate cortex removal as adults. These animals had been behaviorally tested for two years after lesioning and had demonstrated significant recovery of pattern vision. Light and electron microscopy and autoradiography were done on the central retino-recipient nuclei following a monocular intravitreal injection of 3H-proline. Light microscopic analysis of retinal ganglion cell number showed a 30% loss in the parafoveal retina due to retrograde trans-synaptic degeneration. The most striking central change in retinal axon distribution was in the dorsal lateral geniculate nucleus (dLGN) where the parvocellular but not the magnocellular region showed a marked reduction in retinal input. Despite the loss of almost all dLGN neurons through retrograde degeneration, at the EM level both parvocellular and magnocellular regions contained islands of neuropil made up of retinal and several other types of synaptic terminals as well as small dendrites and pale unidentified processes. Approximately equal numbers of retinal terminals were identified by EM autoradiography in both regions of dLGN, which did not explain the apparent differences in labeling between the two regions seen in the light microscope. A second change in central retinal pathways was found in the olivary pretectal nucleus where a significant loss of retinal input also occurred. A third change, but in the opposite direction, was found in the pregeniculate nucleus (PGN) where the area of retinal terminals appeared enlarged. The remaining central retino-recipient nuclei had the same distribution of retinal input as the control animals.

Concordance of inhalation rCBFs with clinical evidence of cerebral ischemia.

Using the 133-Xenon inhalation technique, cerebral blood flow (CBF) and hemispheric blood flow (HBF) were determined serially in 45 patients with acute stroke undergoing pharmacologic trials and in 8 transient ischemic attacks (TIA) schedules for superficial temporal-middle cerebral artery anastomoses. Both patient populations had lower blood flow than a control group of similar ages. Patients in both populations with lateralized clinical signs demonstrated an asymmetry in HBF which corresponded to their clinical signs. In the stroke population, the trend we expected over time toward development of asymmetrical HBF as the non-infarcted hemisphere recovered from diaschisis did not appear.

The primate visual system after bilateral removal of striate cortex. Survival of complex pattern vision.

This study examined the strategies used by monkeys lacking striate cortex to perform visual pattern discriminations. Complete bilateral removal of area 17 initially produced severe visual impairment with recovery of even rudimentary visual capacities (e. g., flux discrimination) dependent on gradually retraining the monkeys through a set of increasingly more complex pattern discriminations. After extended periods of postoperative testing, however, three of five monkeys lacking striate cortex were able to discriminate a number of complex visual patterns even when such local stimulus cues as amount of contour and number of elements were equal. Further testing demonstrated that these animals could distinguish a pattern's spatial organization. They were also able to transfer good performance to tasks with novel patterns.

Residual spatial vision in the monkey after removal of striate and preoccipital cortex.

Rheusus monkeys were trained to discriminate the angular velocity of moving stimuli and to reach accurately toward lighted targets. They were able to recover good performance of these tasks after large bilateral lesions that removed, in successive stages, all of striate cortex and almost all of areas OA, OB, and TEO of preoccipital cortex.

Lack of intralaminar sprouting of retinal axons in monkey LGN.

In the dorsal lateral geniculate nucleus (LGN) of the adult cat there is no evidence for translaminar sprouting of retinal axons to fill sites freed of retinal endings from the other eye. We tested the possibility that retinal axons will sprout to fill denervated retinal sites within laminae of the monkey LGN. In 4 monkeys, retinal ganglion cell axons from either the upper or lower half of the retina were destroyed. To maximize the potential for sprouting in the LGN, on one side of the brain the LGN cells to which the remaining retinal axons normally project were removed by ablation of the appropriate portion of the striate cortex. Three months later the eye receiving the retinal lesion was injected with [3H]proline and the retinal projection to the LGN on both sides of the brain was studied using autoradiography. We found no evidence of intralaminar sprouting of retinal axons either in the normal LGN or in the LGN in which the usual targets of retinal axons had been removed.

Cortical blindness after overlapping retinal-striate lesions: a limit to plasticity in the central visual system.

Lesions in cats, rats, and monkeys that spare more than 2% of the optic tract or visual cortex cause trivial deficits on most measures of vision. Overlap in the topography of the visual system may allow the spared remnant to 'see' a wider field of vision than the physiological map predicts. We tested whether monkeys left with only the lower retinal-field portion of their striate map could see with information coming from the upper half of the retina. In 6 rhesus monkeys the ganglion fibers exiting from the lower half of both retinae were cut with a photocoagulator. Later, the portion of area 17 which, according to the electrophysiological map, controls upper retinal vision, was ablated bilaterally. The combined retinal and striate lesions overlapped to include the entire visual field. Together they produced cortical blindness. The monkeys' performance of two pattern and object tasks remained at chance throughout the survival period. A previous study has described considerable sparing of vision after combined optic tract and visual cortex lesions in cats. Differences in the lesion methods and in the anatomy of the cat and monkey visual system may explain the disagreement.