Categorization is a basic mental process that helps individuals distinguish among groups of negative and positive objects, e.g., poisons and nutrients, or predators and prey. Both the inferior temporal cortex (ITC) and the prefrontal cortex (PFC) are thought to play an important role in the formation of perceptual categories. Monkey experiments have suggested that lateral prefrontal cortex (LPFC) participates in learning and processing visual categories. However, in humans category specific visual agnosia follows inferior temporal (IT) cortex but not LPFC damage. We have used a new behavioral approach to show that both normal monkeys and those with bilateral removal of lateral prefrontal cortex learn and generalize perceptual categories of related visual stimuli rapidly without explicit instruction, indicating that LPFC is not essential for perceptual categorization. Learning to generalize sets of visual stimuli into categories is a valuable cognitive ability for everyday exploratory behavior because it allows individuals to estimate whether newly encountered objects or other stimuli are useful or desirable, or threatening, by accessing associations from past experience without having to learn each object from scratch. Single neuronal recording experiments show that lateral prefrontal cortical neurons show category specific selectivity, which have led to the suggestions that this lateral prefrontal cortex (LPFC) might play a critical role in learning and reacting to stimuli according to their category assignments (3), e.g., dogs are friendly and cats less so. However, humans with inferior temporal, but not lateral prefrontal cortex damage, have agnosias for visual categories, leading to the conclusion that inferior temporal, but not lateral prefrontal cortex is necessary for category formation. These different hypotheses expose a gap in our understanding of how category processing occurs in the brain. A critical test of whether LPFC in category learning and interpretation can be carried out in monkeys by comparing the performance normal monkeys to those with LPFC ablations in a task during which the performance of normal monkeys reveals that they learn and generalize objects into visual categories. In the past, monkeys were taught to report their perception by associating different actions with each perceptual category, e.g., look left for category A and right for category B. Because LPFC is likely to be involved in action-selection processes, it is essential to isolate cognitive requirement from executive ones (e.g., action selection) for the critical test. Here we adapt an approach from our earlier work studying incentive values in monkeys. In our task, the monkeys (N=4) are trained to release a lever when a spot changes from red to green, a process that is completed within a few weeks. After this training, a second visual stimulus, a cue, is presented behind the spot throughout each trial. The cue indicates the incentive value of the reward delivered after a correct response to the green spot (e.g. 2 drops of juice immediately or 1 drop after 7 seconds). The monkeys prediction of the upcoming incentive value is reflected in both reaction time and error rate, where an error is either releasing the bar too early or too late. When we used two categories of visual cues (e.g., 20 dogs = high-incentive;20 cats = low-incentive), by the end of the first testing session the error rates for each of the two categories quickly became significantly different To be sure that the monkeys were not learning one category only in an A, not A strategy, we carried out control experiments (N = 3) showing that the monkey easily learned three categories. All four monkeys were then given bilateral removals of lateral prefrontal cortex (LPFC). This large removal of the prefrontal cortex had no effect on either performance acccuracy (error rate) or quickness on either responding to the categories learned prior to the lesions, generalizing to trial unique cats or dogs, or to learning two new categories (cars vs trucks). These results show that lateral prefrontal cortex is not needed to form perceptual categories. This leaves open which brain regions, presumably in the temporal lobe, are necessary to form perceptual categories. In addition these experiments raise the question of what the role of lateral prefrontal cortex is. Humans with LPFC lesions form categories quickly, also, but show great difficulty in changing their motor responses, once they have established a rule. This suggests that LPFC plays an important role in allowing flexibile adjustment of actions after outcome contingencies change, that is,. LPFC lesions lead to impairments in executive control of action. Thus, post-lesion deficits in humans are interpreted as perseverative, a deficit in changing motor responses after a change in a perceptual classification, which is a deficit related to executive control of action. In a strategy similar to that used above, we devised a delayed matching task where match and non-match trials are associated to alternative reward contingencies but do not instruct a differential behavioral response. The monkeys sit in a chair facing a computer screen and start each trial by holding a touch-bar. They are required to release the bar when a red spot (which appears overlaid to the sample image, in the center of the screen) changes its color into green (occurring after a short random interval from the appearance of the test image). Failing to do so results in the immediate end to the trial. A new trial then began after an intertrial interval. Four monkeys abstracted the concept of trial type across the delay and learned in 3-6 days to use it to predict the forthcoming contingency, as inferred from the highly significant different reaction times and percentage of aborted trials in different trial types (3-8 days if a measure analogous to trials-to-criterion was used). Thus, learning to form abstract concepts can occur quickly in monkeys if it emerges spontaneously from exploratory behavior. The monkeys with LPFC lesions did this also. When the reward contingencies were reversed, the normal monkeys changed their behavior after another 3-6 sessions, whereas the monkeys with the LPFC ablations were severely impaired, taking 16-22 sessions to reverse their behavioral patterns. These results taken together support the hypothesis that LPFC does not participate in perceptual concept formation, but plays a very strong role in flexible adjustment of behavior in response to changes in perceptual concepts.
Showing the most recent 10 out of 14 publications
