Inferotemporal Research Articles

Visual information processing from the retina to the prefrontal cortex

Studies on the neural bases of visual information processing were reviewed by tracing connections from the retina to the prefrontal cortex in the monkey; i.e., retina--lateral geniculate body (LGB)--striate cortex--prestriate cortex--inferotemporal (IT) cortex or posterior parietal (PP) cortex--prefrontal cortex. The retinotopy exists up to the level of prestriate cortex, but it is lost in the IT or PP cortex. A general tendency is observed that the later is the stage of information processing, the longer is the latency of cell activity changes, the larger is the size of receptive field, and more complex become trigger features. Small spot stimuli activate retinal ganglion cells or LGB cells while bars or edges with proper orientation are trigger stimuli for striate and some prestriate cells. In restricted prestriate areas, cells which are selective to color or movement are observed. IT cells are activated by complex pattern stimuli, while PP cells are selective to spatial aspects of the stimulus. Some prefrontal cells encode the meaning of the stimulus independent of its physical properties.

Cold lesions in inferotemporal cortex produce reversible deficits in learning and retention of visual discriminations

Cooling devices were implanted bilaterally at four points along the inferotemporal (IT) cortex in three monkeys. Cooling any of the three anterior pairs disrupted performance on a visual memory task (delayed match-to-sample) uniformly over delays of 0, 15, 30, and 45 sec. Cooling the most posterior cry ode had no effect. Cooling any of the cry odes had no effect on retention of a discrimination between horizontal and vertical stripes. With a focus at the second cryode from anterior, there was significant impairment on learning an object discrimination. With the same focus, a deficit was also found in retention of multiple object discriminations, but there was no deficit in retention of multiple black-white patterns. The anterior two cryodes were then interconnected, as were the posterior two cryodes, so that the anterior or posterior half of IT could be cooled. The animals were tested on retention of a mix of black-white patterns and objects. There were greater deficits on objects anteriorly and patterns posteriorly, but this was margin- ally below statistically acceptable levels. The four pairs of probes were each cooled separately during tests of retention of discrimination of monkey faces. A depression of scores was produced by all four cryodes, but more at the third from anterior. We suggest that IT is differentiated in the way it processes stimuli, not in the localization of such processes as attention, perception, and learning.

Detection of masked patterns in monkeys with inferotemporal, striate or dorsolateral frontal lesions

Monkeys with lateral striate (LS) lesions, inferotemporal (IT) lesions or dorsolateral frontal (DLF) lesions were trained to discriminate between two visual patterns before and after the patterns were masked with irrelevant stimuli (rings enclosing the patterns). Compared to the DLF monkeys, the IT monkeys and one monkey with LS lesions were retarded in discriminating between the unmasked patterns. Irrespective of their performance with unmasked patterns, the IT and LS monkeys were impaired in discriminating the masked patterns, compared to the DLF monkeys. Analysis of performance in testing suggests that both LS and IT lesions produce an impairment in selective attention to visual stimuli.

Electrically-induced afterdischarges in the inferotemporal cortex of monkeys: Effects on visual discrimination and discrimination-reversal performance

Afterdischarges induced bilaterally in inferotemporal (IT) cortex, as monitored by EEG recording, impaired subsequent performance on perceptually difficult visual discriminations of dimensionalized stimuli during the period of abnormal IT EEG activity. The defect seemed to relate to control levels of performance and varied somewhat in persistence. Such seizures also markedly impaired reversals of a visual pattern discrimination. Based upon information about propagation patterns of the afterdischarges, it was suggested that afterdischarges induced in IT cortex produce some form of visual perceptual or attentional disorder and an associative disorder in the encoding of visual stimuli with respect to reward contingencies.

The interaction of visual attention and temporal cortex stimulation on electrical activity evoked in the striate cortex

Four experiments were directed at finding an index of the activity of the infero-temporal (IT) cortex on the visual system. In the first experiment, electrical stimulation of the IT cortex failed to alter recovery functions in the striate cortex of four monkeys. This result was different from that reported in an earlier experiment, but comparison of the initial recovery functions obtained in the two experiments allowed us to attribute the discrepancy to differences in the attentional state of the monkeys. This suggestion was tested by developing, in the second experiment, a reliable measure of state in the visual system, a measure responsive to environmental variables inducing attention. This measure was used in the third experiment to select and classify recovery cycles on the basis of the state of the visual system. Only when the attentive state was present were the recovery cycles similar to those recorded in the first experiment. Finally, the measure, a potential evoked in the striate cortex by geniculate stimulation, was shown to be sensitive to IT stimulation only when monkeys were in an inattentive state.

Impairment in sampling visual stimuli in monkeys with inferotemporal lesions

Monkeys with inferotemporal (IT) lesions and their controls, monkeys with partial removal of striate cortex (LS) and unoperated monkeys, were trained to discriminate between two stimulus compounds differing in brightness near the response site and in hue, distant from the response site. In subsequent discrimination testing, only the distant cue (hue) was available. All animals learned the original discrimination rapidly. However, in the discrimination test, the IT monkeys made significantly more errors than did the unoperated monkeys, while the LS monkeys were unimpaired. These findings support the view that IT lesions impair visual search. Once the IT monkeys learned the test discrimination, they were not impaired in a series of subsequent tests in which the area of the distant cue was successively reduced.

Impairments in selective attention to visual stimuli in monkeys with inferotemporal and lateral striate lesions

Bilateral removal of inferotemporal (IT) cortex in monkeys produces a severe impairment in discrimination learning limited to the visual modality (see Mishkin 12 for a review of these findings). While the visual deficits produced by IT lesions cannot be readily attributed to sensory losses 7,16 the nature of the disorder remains unclear. Chow and Orbach 5 attempted to determine whether IT ablation is associated with losses in selective attention by use of a discrimination task in which visual stimuli are briefly exposed, but found no impairment in IT monkeys in this situation. Subsequently, it was shown that the failure to obtain visual losses in IT monkeys under these conditions could be attributed to preoperative overtraining 6,13. The present experiment reinvestigated the question of selective attention losses in IT monkeys by methods which avoid the problem of preoperative overtraining. In this experiment, monkeys with IT ablation, along with their controls monkeys with lateral striate (LS) ablation and unoperated monkeys were first trained to discriminate between two visual stimuli. Irrelevant features were then added to the stimuli in a series of tests in which the animals were required to relearn the discrimination problem. If the IT animals were to be retarded in relearning the discrimination, this might imply a loss in the ability to selectively attend to or abstract the relevant visual features.

The effect of discrimination training on pattern equivalence in monkeys with inferotemporal and lateral striate lesions

Monkeys with inferotemporal (IT) lesions, monkeys with lateral striate (LS) lesions and unoperated controls discriminated between two patterns and were tested for pattern equivalence in three experiments. The IT monkeys appeared to utilize salient cues more than control animals did and failed to identify specific pattern features (exps. 1 and 2). When the salient cues were removed (exp. 3) the IT animals gave evidence of normally identifying configurational cues. While two LS monkeys showed deficits like those of the IT monkeys (exps. 1 and 2), they apparently did not identify configurational features (exp. 3). The findings suggest that IT lesions do not impair the capacity to identify visual features, but disrupt stimulus sampling or attentional processes.

Size Discrimination and Transposition in Monkeys with Striate and Temporal Lesions

Monkeys with inferotemporal (IT) lesions, lateral striate (LS) lesions and unoperated controls were trained to discriminate between two stimuli differing in size and were then tested for size transposition to stimuli altered in absolute size, in shape and in linear dimensions (height and width). The LS animals were more severely impaired than were the IT animals in learning the discrimination. In transposition testing the LS monkeys, and to a lesser extent the IT monkeys, differed from the N monkeys in that they showed better size transposition. It is suggested that this paradoxical enhancement of transposition following cortical lesions might be due to an impairment in detecting differences between the training and test stimuli.