Pattern Discrimination Learning Research Articles

Pattern discrimination thresholds after partial inferior temporal or lateral striate lesions in monkeys

Ablation of inferior temporal (IT) cortex, particularly of the posterior region, produces severe impairment in pattern discrimination learning. The present study examined whether this impairment is associated with raised pattern discrimination thresholds. Groups of three monkeys each were given either anterior IT, posterior IT, or foveal striate lesions, or kept as controls. They were trained after surgery on a threshold task in which a 90° white angle on a gray ground was the standard, and 15 angles ranging from 10° through 88.5° were the comparisons. As expected, monkeys with posterior IT lesions were the most severely impaired in learning the initial discrimination (90° vs. 10°). However, only the monkeys with foveal striate lesions showed significant impairment on the subsequent threshold determinations. The results indicate that raised pattern discrimination thresholds are not the cause of the pattern discrimination learning deficits produced by inferior temporal lesions. Data from additional visual discriminations presented after threshold testing was completed point, instead, to a loss of attention to stimulus features as the explanation for the learning deficit.

The effects of instability of the visual display on pattern discrimination learning by monkeys: Dissociation produced after resections of frontal and inferotemporal cortex

Rhesus monkeys with anterior frontal ( N=7), posterior parietal ( N=4), or inferotemporal ( N=4) lesions and normal controls ( N=3) were trained on visual discrimination problems under conditions in which the stimulus display either remained in the same place or shifted randomly over a wide range of positions from trial to trial. Naïve monkeys with frontal lesions were significantly impaired in learning the discrimination problem under the condition of shifting spatial context; their impairment disappeared with sophistication, but they continued to find this condition more difficult than one in which the display remained stable. The performance of monkeys with inferotemporal lesions was markedly improved by the random-position display condition.

Some variables in auditory pattern discrimination learning

Twelve rhesus monkeys learned auditory frequency and pattern discriminations more rapidly than 6 cats in the same conditioned avoidance situation. The same 12 monkeys learned the same frequency and pattern discriminations much more rapidly than 4 monkeys trained under conditions of orange juice reinforcement. Five additional monkeys learned the same auditory pattern discrimination as that learned by the 12 monkeys in fewer trials when the pattern task was one of 3 or 4 auditory pattern discriminations learned simultaneously. The data from the multiple pattern monkeys offer insight into the role of pattern structure in the rate of pattern discrimination learning.

Some variables in the learning of auditory pattern discriminations by animals

We have examined the role of a number of variables in auditory pattern discrimination learning: (1) species, (2) reinforcement, (3) learning sequence, and (4) internal structure of the auditory patterns. (1) In a conditioned avoidance situation cats learned a simple frequency discrimination more slowly than rhesus monkeys—and an auditory pattern discrimination at about the same rate as monkeys. (2) Four monkeys working for juice reinforcement required a mean number of 737 trials to learn the same frequency discrimination as 12 monkeys working to avoid shock learned in a mean of 55 trials. The same four monkeys learned an auditory pattern task in a mean of 451 trials for juice, while 12 monkeys in the shock avoidance situation required a mean of 159 trials. (3) Five monkeys learning three or four auditory pattern discriminations simultaneously required a mean of 11 trials to learn the same auditory pattern discrimination for which 12 monkeys required a mean of 159 trials to learn. (4) Five monkeys learned to differentiate patterns which consist of different proportions of the same frequency elements more rapidly than they learned to discriminate patterns with the same proportion of frequency elements but differently arranged. These results are discussed on the background of earlier studies of auditory pattern learning by animals. [Supported by PHS NS 06587.]

Modifications of pulvinar and geniculo-cortical evoked potentials during visual discrimination learning in monkeys

Electrical activity was recorded from various cortical and subcortical loci in monkeys during the learning of visual pattern discriminations. Potentials evoked by the discriminative stimuli were found to change during learning in the striate and prestriate cortex, the lateral geniculate body and the inferior pulvinar. No changes were evident in the medial pulvinar, midbrain reticular formation or hippocampus. In the inferior pulvinar and prestriate cortex the greatest changes were evident during the acquisition of the discrimination, while in the lateral geniculate nucleus and striate cortex the most marked changes were observed following attainment of criterion performance. These results are interpreted as showing a functional differentiation of the pulvinar-prestriate system from the geniculostriate system, and suggest that the inferior pulvinar is involved in the acquisition of visual pattern discriminations.

Effects of neonatal and adult striate lesions on visual discrimination in the rabbit

Learning of brightness and pattern discriminations was compared in normal rabbits and in rabbits with striate lesions performed in infancy or adulthood. Those with lesions did not differ from normal rabbits in learning a brightness discrimination. Both groups with lesions took significantly longer than normal to learn patterns consisting of striations differeing by 90° and 45°. No animal with lesions learned to discriminate striations differing by 22°. There were no significant differences between animals with lesions performed in adulthood and infancy in learning brightness or pattern discriminations. Species differences in neonatal plasticity and functional recovery are discussed.

Unilateral pleasure-center stimulation in split-brain cats

Unilateral bipolar stimulation at seven positively reinforcing sites ranging from the ventral mesencephalic reticular formation to the preoptic region, in five split-brain cats, was shown to reinforce pattern discrimination learning in both hemispheres. The training method used permitted simultaneous presentation of discriminanda to both hemispheres of the split-brain animal. The separated hemispheres were subsequently tested individually for establishment of learning, and it was found that learning had occurred in both hemispheres, in most cases. Monocular training showed reinforcement of behavior can be produced in both hemispheres at every electrode site tested. It is concluded that unilateral pleasure-center stimulation may exert a bilateral central effect via direct and crossed projections from the brain stem.

Lateralization of brightness and pattern discrimination learning in the corpus callosum-sectioned rat

Lateralization of brightness and pattern discrimination learning in the corpus callosum-sectioned rat

Pattern versus component discrimination learning with extended training.

Pattern versus component discrimination learning with extended training.

The significance of crossed and uncrossed projections for the acquisition and retrieval of visual engrams in hooded and albino rats.

Occlusion of one eye and functional hemidecortication by cortical spreading depression (CSD) were used to examine the relative importance of crossed and uncrossed visual projections for pattern discrimination learning in hooded and albino rats. Whereas functionally hemidecorticate albino rats were unable to learn in 200 trials a horizontal-vertical discrimination when using the eye contralateral to the depressed hemisphere, hooded rats needed under similar conditions three times as many trials to criterion (167) than when using opposite eye and hemisphere (59). An engram established in intact animals with both eyes could be retrieved in functionally hemidecorticate rats of both strains even with the eye ipsilateral to the intact hemicortex, but retraining required about twice as many trials as when using the contralateral eye. Retrieval of monocularly acquired engrams was less efficient under similar conditions. Pattern discrimination formed through crossed or uncrossed visual projections during monocular occlusion and contralateral or ipsilateral CSD was completely lateralized, since relearning with naive eye and hemisphere proceeded at the same rate as the original learning. On the contrary, a black-white discrimination trained with the eye contralateral to the depressed hemisphere could be partly retrieved by the untrained eye and hemisphere. It is concluded that the ipsilateral visual projection is more effective in hooded than in albino rats and that it can be more efficiently used for retrieval than for acquisition.

Different effects of temporal lobe lesions on learning of visual pattern discriminations and of a visuomotor coordination task

The experiment described was designed to test if learning of a visual guidance response was impaired as severely as pattern-discrimination learning by lesions of the inferior temporal neocortex. Two groups of split-brain monkeys were used. One group, the experimental animals, also sustained unilateral removals of the inferior temporal area and the hippocampus. The other group, the control animals, did not receive further lesions. In addition two animals of the control group later received the unilateral lesions and were subsequently retrained on the task. The animals were taught to retrieve a peanut from a revolving container using only one eye and the contralateral hand. The time taken for the retrieval was measured. The eye and hand used for training were changed each day. Five sessions with each eye were given. It was found that the majority, but not all. of the split-brain hemispheres were more proficient on the final session than on the first, indicating that learning had taken place. It was also found that the learning for the two hemispheres did not differ, except for isolated sessions. Furthermore the level reached by the animals of the control group that later received lesions was not lost following the damage. It is concluded that the learning seen in this experiment does not depend on the temporal lobe structures in the same way as pattern discrimination learning.

Errata: Object and Pattern Discrimination Learning by Young Children as a Function of Availability of Cues

Errata: Object and Pattern Discrimination Learning by Young Children as a Function of Availability of Cues

An investigation of a possible memory defect produced by inferotemporal lesions in the baboon

Bilateral inferotemporal lesions and more extensive lesions which include medial temporal damage severely impair visual pattern and object discrimination learning in the baboon. Furthermore, such animals show poor retention of simple object discrimination problems learned 24 hours before and continue to show poor retention if thesame problems are relearned for the second time after a further 24 hour delay. The basis of this learning and memory deficit in the inferotemporals is discussed in the context of experiments showing impaired perceptual analysing mechanisms after such lesions and of more recent studies indicating separate foci for visual perception and memory in the ventral temporal lobe cortex. An effort is made to reconcile the spectrum of deficits seen in the animals which sustained additional medial temporal damage with the dramatic memory impairments seen after bilateral medial temporal lobe damage in man.

Pattern Area and Discrimination Learning with Pattern Spacing, Pattern Height, Pattern-Border Contiguity, and Pattern Manipulation Controlled

The relationship between pattern area and pattern discrimination learning was explored. The possibly confounding effects of pattern spacing, height, contiguity, and manipulation were removed or controlled by: (a) presenting illuminated .25-, 1-, and 4-sq. in. patterns in a totally dark room, (b) alternating correct and incorrect patterns in a randomly determined temporal succession, (c) equating pattern height for all pattern-area conditions, (d) prohibiting pattern manipulation. Increasing pattern height from 1 to 4 sq. in. still resulted in improved learning. Results were interpreted in terms of utilization of brightness and/or size differences as cues to pattern discrimination.

Enhanced pattern discrimination learning following unilateral damage to posterior cortex in rats

Rats with unilateral damage to the posterior cortex, including the “visual” cortex, learned a monocular pattern discrimination substantially more quickly than did an unoperated group. A unilaterally blinded group did not differ reliably from unoperated animals.

Stimulus information and primate pattern discrimination: Simultaneous changes in positive and negative cue information

The present research studied the effects for two-choice, discrete-trial, pattern-discrimination learning of changes in the stimulus-information content of both the positive and negative stimulus cues. The major findings suggest that, with experienced primates, there exists a relation between equal increments in positive and negative stimulus information and discrimination learning that is strikingly similar to the classical psychophysical relationship between stimulus intensity and JNDs.

Object and pattern discrimination learning by young children as a function of availability of cues.

Children of CA 4 and 6 years were presented with stereometric, planometric, or pattern stimuli in a 3-choice discrimination problem. Half (allcues group) were unrestricted with regard to availability of cues, and half (visual-cues group) were restricted to the use of visual cues. The all-cues group learned to discriminate stereometric objects more rapidly than the visual-cues group. Planometric and pattern discriminations were learned equally rapidly by all-cues and visual-cues groups. 6-year-olds performed equally well and significantly above chance level when trained on all three types of discrimination. 4-year-olds performed above chance level only when trained with stereometric objects. The middle-sized stimulus was the most difficult to discriminate for all Ss.

Object and Pattern Discrimination Learning by Young Children as a Function of Availability of Cues

Object and Pattern Discrimination Learning by Young Children as a Function of Availability of Cues

A comparison of the effects of inferotemporal and striate cortex lesions on the visual behaviour of rhesus monkeys.

The effects of bilateral removal of inferotemporal cortex or lateral striate cortex are compared. The former operation impairs visual pattern discrimination learning, without disturbing prompt detection and retrieval of food, visual acuity or visual fields. In contrast, animals in which the macular striate projection area has been removed are significantly superior in tests of visual pattern discrimination learning despite showing impairment of visually guided reaching, visual acuity, and visual fields. The results indicate that the inferotemporal defect is not caused by deficient visual sensitivity or acuity.

Pattern discrimination learning with rhesus monkeys.

Rhesus monkeys were trained to criterion on a 2-stimulus and a 5-stimulus pattern discrimination task. The probabilities of response to the various stimuli throughout learning are examined for individual Ss, and it is found that Ss exhibit consistency in the order and manner in which incorrect stimuli are eliminated. This suggests a simple mathematical description of the process, which is used to deepen the analysis of the data, permitting estimation of individual learning parameters and construction of more meaningful summaries of the group data.