Pattern Discrimination Learning Research Articles

Effects of MT lesions on visuomotor performance in macaques.

Monkeys with selective bilateral lesions of area MT were trained on tasks designed to examine visuomotor function. They were required to: 1- retrieve a small food pellet from a narrow slot; 2- locate and retrieve a loose peanut mounted on a background of fixed peanuts; and 3- retrieve an erratically moving food pellet from a spinning bowl. After the lesions, these monkeys were behaviorally impaired relative to their own preoperative performances and also relative to the postoperative performances of the control monkeys with lesions in optic radiation fibers (OR) under MT or lesions in the posterior parietal cortex (PP). Although their performance improved with practice and time, the MT-lesioned monkeys showed long-term impairments twenty weeks after surgery. Control monkeys performed no worse on the tasks after their lesions. Another task which required the monkeys to retrieve a food pellet without visual guidance revealed that all the animals performed equally poorly when visual cues were unavailable, but that only the control monkeys benefited when visual cues were available. None of the monkeys were impaired on a pattern discrimination learning task. Besides that, direct observations revealed that the MT-lesioned animals grasped peanuts in a manner different from the control animals.

Genetic inactivation of hypoxia inducible factor 1-alpha (HIF-1α) in adult hippocampal progenitors impairs neurogenesis and pattern discrimination learning.

HIF-1α is a hypoxia-inducible protein that regulates many cellular processes, including neural stem cell maintenance. Previous work demonstrated constitutive stabilization of HIF-1α in neural stem cells (NSCs) of the adult mouse subventricular zone (SVZ) and hippocampal subgranular zone (SGZ). Genetic inactivation of NSC-encoded HIF-1α in the adult SVZ results in gradual loss of NSCs, but whether HIF-1α is required for the maintenance of SGZ hippocampal progenitors and adult hippocampal neurogenesis has not been determined. Here we tested the hypothesis that HIF-1α plays an essential role in the maintenance of adult hippocampal neurogenesis using Nestin-CreERT2/R26R-YFP/Hif1afl/fl triple transgenic mice, in which HIF-1α was genetically inactivated in nestin+ hippocampal progenitors and their downstream progeny following tamoxifen exposure. We found that disruption of HIF-1α gene expression resulted in a marked 50% reduction of adult-generated dentate granule cells (DGCs) that was highly correlated with impaired hippocampal function, as assessed using two behavioral assays of pattern discrimination. These behavioral tests included the A-B contextual fear-conditioning task and the trial-unique, delayed nonmatching-to-location (TUNL) touch-screen operant chamber task. Our findings identify HIF-1α as a novel regulator of adult hippocampal neurogenesis under non-pathological conditions, and underscore the importance of neurogenesis for pattern discrimination learning.

Genetic inactivation of synaptosomal-associated protein 25 (SNAP-25) in adult hippocampal neural progenitors impairs pattern discrimination learning but not survival or structural maturation of newborn dentate granule cells.

Adult neurogenesis is necessary for proper cognition and behavior, however, the mechanisms that underlie the integration and maturation of newborn neurons into the pre-existing hippocampal circuit are not entirely known. In this study, we sought to determine the role of action potential (AP)-dependent synaptic transmission by adult-generated dentate granule cells (DGCs) in their survival and function within the existing circuitry. We used a triple transgenic mouse (NestinCreERT2 :Snap25fl/fl : tdTomato) to inducibly inactivate AP-dependent synaptic transmission within adult hippocampal progenitors and their progeny. Behavioral testing in a hippocampal-dependent A/B contextual fear-discrimination task revealed impaired discrimination learning in mice harboring SNAP-25-deficient adult-generated dentate granule cells (DGCs). Despite poor performance on this neurogenesis-dependent task, the production and survival of newborn DGCs was quantitatively unaltered in tamoxifen-treated NestinCreERT2 :Snap25fl/fl : tdTomato SNAP compared to tamoxifen-treated NestinCreERT2 :Snap25wt/wt : tdTomato control mice. Although SNAP-25-deficient adult DGCs displayed a small but statistically significant enhancement in proximal dendritic branching, their overall dendritic length and distal branching complexity was unchanged. SNAP-25-deficient newborn DGCs also displayed robust efferent mossy fiber output to CA3, with normal linear density of large mossy fiber terminals (LMTs). These studies suggest that AP-dependent neurotransmitter release by newborn DGCs is not essential for their survival or rudimentary structural maturation within the adult hippocampus.

Multi-layer network utilizing rewarded spike time dependent plasticity to learn a foraging task.

Neural networks with a single plastic layer employing reward modulated spike time dependent plasticity (STDP) are capable of learning simple foraging tasks. Here we demonstrate advanced pattern discrimination and continuous learning in a network of spiking neurons with multiple plastic layers. The network utilized both reward modulated and non-reward modulated STDP and implemented multiple mechanisms for homeostatic regulation of synaptic efficacy, including heterosynaptic plasticity, gain control, output balancing, activity normalization of rewarded STDP and hard limits on synaptic strength. We found that addition of a hidden layer of neurons employing non-rewarded STDP created neurons that responded to the specific combinations of inputs and thus performed basic classification of the input patterns. When combined with a following layer of neurons implementing rewarded STDP, the network was able to learn, despite the absence of labeled training data, discrimination between rewarding patterns and the patterns designated as punishing. Synaptic noise allowed for trial-and-error learning that helped to identify the goal-oriented strategies which were effective in task solving. The study predicts a critical set of properties of the spiking neuronal network with STDP that was sufficient to solve a complex foraging task involving pattern classification and decision making.

Generalized Perceptual Learning in the Absence of Sensory Adaptation

Repeated performance of visual tasks leads to long-lasting increased sensitivity to the trained stimulus, a phenomenon termed perceptual learning. A ubiquitous property of visual learning is specificity: performance improvement obtained during training applies only for the trained stimulus features, which are thought to be encoded in sensory brain regions [1-3]. However, recent results show performance decrements with an increasing number of trials within a training session [4, 5]. This selective sensitivity reduction is thought to arise due to sensory adaptation [5, 6]. Here we show, using the standard texture discrimination task [7], that location specificity is a consequence of sensory adaptation; that is, it results from selective reduced sensitivity due to repeated stimulation. Observers practiced the texture task with the target presented at a fixed location within a background texture. To remove adaptation, we added task-irrelevant ("dummy") trials with the texture oriented 45° relative to the target's orientation, known to counteract adaptation [8]. The results indicate location specificity with the standard paradigm, but complete generalization to a new location when adaptation is removed. We suggest that adaptation interferes with invariant pattern-discrimination learning by inducing network-dependent changes in local visual representations.

Learning and recall of object and pattern discriminations during bilateral reversible deactivation of the superior colliculus

Learning and recall of object and pattern discriminations during bilateral reversible deactivation of the superior colliculus

Learning to see the trees before the forest: Reversible deactivation of the superior colliculus during learning of local and global visual features

Previous studies have established that deactivation of the superior colliculus severely retards the normally rapid learning of pattern discriminations in the mature cat. The purpose of this study was to test the hypothesis that the midbrain plays an important role in the learning of simple pattern discriminations and that the contribution of this pathway is to the perception of global, rather than local, features of a figure. To answer this question, pattern discrimination learning was studied in three intact cats and in three experimental cats during bilateral reversible deactivation of the superficial layers of the superior colliculus (SC). The animals concurrently learned to discriminate three pairs of compound visual patterns composed of small (local element) Ts or 7s. Congruent and incongruent stimulus pairs were large (global) Ts vs. 7s comprising the same or different local elements, respectively. The random stimulus pair consisted of randomly scattered local elements (Ts vs. 7s). The animals were trained to respond to the large Ts in the congruent and incongruent pairs and the small Ts in the random pair. In the normal cats, learning of the random pair was much slower than the learning of the congruent and incongruent pairs. This finding demonstrated the theory of global precedence, because the animals learned the global features of the congruent and incongruent pairs much more quickly than the local features of the random pair. In contrast, during bilateral deactivation of the superficial layers of the SC, the learning of the incongruent pair was significantly retarded and took longer to learn than the random pair. Congruent and random pair learning rates were unchanged. The specific deficit in learning the incongruent pair indicates that the learning of global, but not local, elements of the visual pattern is impaired during deactivation of the SC. The unimpaired use of local features permitted the animals to learn the congruent and random pairs at normal rates. Therefore, deactivation of the superficial layers of the SC during pattern discrimination learning reverses the precedence for global visual features that is typical of normal learning.

Learning to see the trees before the forest: reversible deactivation of the superior colliculus during learning of local and global visual features.

Previous studies have established that deactivation of the superior colliculus severely retards the normally rapid learning of pattern discriminations in the mature cat. The purpose of this study was to test the hypothesis that the midbrain plays an important role in the learning of simple pattern discriminations and that the contribution of this pathway is to the perception of global, rather than local, features of a figure. To answer this question, pattern discrimination learning was studied in three intact cats and in three experimental cats during bilateral reversible deactivation of the superficial layers of the superior colliculus (SC). The animals concurrently learned to discriminate three pairs of compound visual patterns composed of small (local element) Ts or 7s. Congruent and incongruent stimulus pairs were large (global) Ts vs. 7s comprising the same or different local elements, respectively. The random stimulus pair consisted of randomly scattered local elements (Ts vs. 7s). The animals were trained to respond to the large Ts in the congruent and incongruent pairs and the small Ts in the random pair. In the normal cats, learning of the random pair was much slower than the learning of the congruent and incongruent pairs. This finding demonstrated the theory of global precedence, because the animals learned the global features of the congruent and incongruent pairs much more quickly than the local features of the random pair. In contrast, during bilateral deactivation of the superficial layers of the SC, the learning of the incongruent pair was significantly retarded and took longer to learn than the random pair. Congruent and random pair learning rates were unchanged. The specific deficit in learning the incongruent pair indicates that the learning of global, but not local, elements of the visual pattern is impaired during deactivation of the SC. The unimpaired use of local features permitted the animals to learn the congruent and random pairs at normal rates. Therefore, deactivation of the superficial layers of the SC during pattern discrimination learning reverses the precedence for global visual features that is typical of normal learning.

Contrasting Effects on Discrimination Learning after Hippocampal Lesions and Conjoint Hippocampal–Caudate Lesions in Monkeys

Eighteen monkeys with lesions of the hippocampal region (the hippocampus proper, the dentate gyrus, and the subiculum) made by an ischemic procedure, radio frequency, or ibotenic acid were tested on a simple, two-choice object discrimination learning task that has been shown to be sensitive to large lesions of the medial temporal lobe. The monkeys were also tested on two other discrimination tasks (pattern discrimination and eight-pair concurrent discrimination) that can be learned normally by monkeys with large medial temporal lobe lesions. All of the lesion groups were impaired at learning the simple object discrimination task. Seven of the monkeys who had sustained damage to the hippocampal region also sustained damage to the tail of the caudate nucleus. These seven monkeys, but not the other 11 monkeys with hippocampal lesions, were impaired on pattern discrimination and concurrent discrimination learning. The results suggest that the hippocampal region is important for learning easy, two-choice discriminations, whereas the caudate nucleus is necessary for the normal learning of more difficult, gradually acquired discrimination tasks. The findings support the distinction between declarative memory, which depends on the hippocampus and related medial temporal lobe structures, and habit learning, which depends on the caudate nucleus.

Protective effects of fetal neocortical transplants on cognitive function and neuron size in rats with congenital micrencephaly

The rat with micrencephaly, produced by prenatal exposure to methylazoxymethanol, provides a useful model to study neurobehavioral abnormalities associated with congenital brain defects. The micrencephalic animals have a life-long learning impairment. As they age, their already impaired learning competence deteriorates further. To determine whether the aging-associated functional deterioration could be ameliorated by a neural transplant, micrencephalic rats bearing solid transplants of normal fetal neocortical tissue since infancy were evaluated on a visual pattern discrimination learning at 15 months and a spatial navigation test at 24 months of age. The transplant-bearing rats learned both tasks significantly better than the micrencephalic rats without transplants. Morphometric analyses revealed that cortical pyramidal neurons were larger in the transplant-bearing rats than in micrencephalic rats without transplants. The life-long presence of a transplant appeared to have protected the micrencephalic brain against aging-associated deterioration. This is the first demonstration that a neural transplant, placed in a congenitally defective infant brain, can ameliorate aging-associated cognitive deficits.

An Automated Training Device for Pattern Discrimination Learning of Group-Housed Gerbils

Meier, M., R. Reinermann, J. Warlich and G. Manteuffel. An automated training device for pattern discrimination learning of group-housed gerbils. Physiol Behav 63(4) 497–498, 1998. The setup, designed for the rodent Meriones unguiculatus (gerbil), allows flexible stimulus presentations and rewarding as well as on-line data registration. It consists of a spacious housing where the animals have free access to water. Food is supplied exclusively in the y-maze training compartment and serves as a reinforcer in an operant conditioning paradigm.

The role of visual field position in pattern-discrimination learning.

Invariance of object recognition to translation in the visual field is a fundamental property of human pattern vision. In three experiments we investigated this capability by training subjects to distinguish between random checkerboard stimuli. We show that the improvement of discrimination performance does not transfer across the visual field if learning is restricted to a particular location in the retinal image. Accuracy after retinal translation shows no sign of decay over time and remains at the same level it had at the beginning of the training. It is suggested that in two-dimensional translation invariance-as in three-dimensional rotation invariance-the human visual system is relying on memory-intensive rather than computation-intensive processes. Multiple position- and stimulus-specific learning events may be required before recognition is independent of retinal location.

An artificial neural network analogue of learning in autism

An artificial neural network is simulated that shares formal qualitative with the selective attention and generalization deficits seen in people with autism.The model is based on neruopathological studies which suggest that affected individuals have either too few or too many neuronal connections in various regions of the brain. In simulations where the model was taught to discriminate children with autism from children with mental retardation, having too few simulated neuronal connections led to relatively inferior discrimination of the two groups in a training set and, consequently, relatively inferior generalization of the discrimination to a novel test set. Too many connections produced excellent discrimination but inferior generalization because of overemphasis on details unique to the training set. It is concluded that, within the context of the current model, the neuropathological observations that have been described in the literature are sufficient to explain some of the unique pattern recognition and discrimination learning abilities seen in some people with autism as well as their problems with generalization and concept acquisition. The model generates testable hypothesis that have implications for understanding the pathogenesis, treatment, and phenomenology of autism.

AMNESIA FOLLOWING DAMAGE TO THE LEFT FORNIX AND TO OTHER SITES: A COMPARATIVE STUDY

Two memory-impaired patients, who had suffered damage to the left or both fornix columns during removal of a ventricular cyst, were compared with 3 others having left-sided hippocampal or thalamic lesions, and with normal controls. The tests used were nonverbal--scene recognition, delayed matching-to-sample and concurrent pattern and object discrimination learning. The last two are differentially sensitive to fornix transection and to hippocampal or thalamic ablations in monkeys; however, the patients with fornix damage did not show a distinctive pattern of impairment. The reasons for this discrepancy are discussed. The study adds to the evidence that fornix transection can cause wide-ranging memory disturbances in man.

Concurrent and Sequential Pattern Discrimination Learning by Patients with Korsakoff Amnesia

Patients with Korsakoff's syndrome and alcoholic controls learned to discriminate sets of pairs of patterns presented concurrently, in order to test predictions based on monkeys' performance in similar tasks following medial temporal or diencephalic lesions. In Experiment 1 the subjects learned a 2-pair, a 6-pair and a 10-pair set; the Korsakoff group were impaired on the first and last, but not on the second set. In Experiment 2, the same subjects learned single pairs sequentially, and 2-pair and 8-pair sets concurrently. The effect of 2 types of feedback for correct responses (visual or non-visual) was also compared. The Korsakoff patients were markedly poorer than controls under all conditions; the type of feedback made little difference. In several respects the patients' impairment differed from what had been predicted from the animal experiments.

Two-Stage Model of Visual Pattern Discrimination Learning in Macaque Monkeys (Macaca mulatta and Macaca fuscata)

To investigate the nature underlying the process of pattern discrimination learning, a series of seven experiments on seven working assumptions were undertaken. The main findings are as follows. The pattern discrimination learning consists of two time-dependent stages: The initial or first learning stage is the period of performance at chance, and the succeeding or second stage is the period of performance from just above the chance to a criterion level. The duration of the first stage is dependent on the degrees of cue-response separations, whereas that of the second stage is independent of. During the first stage, monkeys do not attend to the discriminative cue even at small cue-response separations, whereas during the second stage, they achieve pattern discrimination, or pattern perception and cognition, regardless of cue-response separations. After having learned the first pattern task with a cue-response separation, they learned new pattern tasks by means of the second stage, showing marked saving of the duration of the first stage. The findings in the present studies indicate that the first stage of learning is predominantly involved in the process of attending to the discriminative cues remote from the response site (a selective attention to the cue), whereas the second stage is concerned with the process of perception and cognition of the discriminative cue.

Partition between stimuli slows down greatly discrimination learning in binocularly deprived cats

Pattern discrimination learning was compared in cats reared in a normal environment during the first months of life (N cats), control cats reared in the laboratory with opened eyes (C cats), and cats deprived binocularly of patterned vision from birth (BD cats). A two-choice apparatus was used for training. In Expt. I the cats had been trained without a partition between stimuli and retrained with the partition. In Expt. II the partition was used from the beginning of training. The partition slowed down the learning in all groups. However, the BD cats were much more affected than N cats, and C cats seemed to be more affected than N cats and less than BD cats. In BD and C cats the number of errors of entering an incorrect alley without pushing the gate was increased. In contrast, when the partition was absent and discrimination very simple (stimulus vs no stimulus), the BD cats seemed to master the task even more quickly than N and C cats. The results suggest that the deficit produced by the partition consisted largely in a difficulty in rapid choosing a correct directional response at distance. In BD cats the mechanism of the deficit was complex.

The effects of anteromedial cortex lesions on pattern discrimination learning in rat

The visual angular difference threshold of rats with unilateral anteromedial cortex lesions was measured. The stereotyped running pattern and preoperatively preferred side of approach in a two-choice visual discrimination apparatus was recorded for each animal. Lesions in the hemisphere ipsilateral to the preferred approach side did not affect either the running pattern or the angular difference threshold. However, there was a general reduction in the detection of the positive stimulus throughout the scale. Lesions in the contralateral hemisphere produced a transient reversal of the side preference with an overall loss of S + detection in the field contralateral to the side approached.

Monocular pattern discrimination learning after unilateral lesions of anteromedial cortex in rat

Monocularly patched rats were trained on a series of visual tilt discriminations ordered from simple to difficult to measure the angular difference threshold of each eye. Preoperatively there was no difference between discrimination performance for either eye. Unilateral lesions of anteromedial cortex did not alter the monocular angular difference threshold. However, there was a reduction in discrimination performance for all angular difference comparisons. Animals showed a visual neglect for stimuli in the field contralateral to the lesion. The neglect was independent of the eye used for the discrimination.

Visual pattern discrimination learning in binocularly deprived cats

Visual pattern discrimination learning in binocularly deprived cats