Biomedical subjects

S Dehaene

Publications and source records attributed to S Dehaene.

60 records · Page 4Linked to original sources

PubMed1989

Discriminability and dimensionality effects in visual search for featural conjunctions: a functional pop-out.

Treisman and Gelade's (1980) feature-integration theory of attention states that a scene must be serially scanned before the objects in it can be accurately perceived. Is serial scanning compatible with the speed observed in the perception of real-world scenes? Most real scenes consist of many more dimensions (color, size, shape, depth, etc.) than those generally found in search paradigms. Furthermore, real objects differ from each other along many of these dimensions. The present experiment assessed the influence of the total number of dimensions and target/distractor discriminability (the number of dimensions that suffice to separate a target from distractors) on search times for a conjunction of features. Search was always found to be serial. However, for the most discriminable targets, search rate was so fast that search times were in the same range as pop-out detection times. Apparently, greater discriminability enables subjects to direct attention at a faster rate and at only a fraction of the items in a scene.

Adult↗

PubMed1987

Neural networks that learn temporal sequences by selection.

A model for formal neural networks that learn temporal sequences by selection is proposed on the basis of observations on the acquisition of song by birds, on sequence-detecting neurons, and on allosteric receptors. The model relies on hypothetical elementary devices made up of three neurons, the synaptic triads, which yield short-term modification of synaptic efficacy through heterosynaptic interactions, and on a local Hebbian learning rule. The functional units postulated are mutually inhibiting clusters of synergic neurons and bundles of synapses. Networks formalized on this basis display capacities for passive recognition and for production of temporal sequences that may include repetitions. Introduction of the learning rule leads to the differentiation of sequence-detecting neurons and to the stabilization of ongoing temporal sequences. A network architecture composed of three layers of neuronal clusters is shown to exhibit active recognition and learning of time sequences by selection: the network spontaneously produces prerepresentations that are selected according to their resonance with the input percepts. Predictions of the model are discussed.

Animals↗

PubMed1986

Spin glass model of learning by selection.

A model of learning by selection is described at the level of neuronal networks. It is formally related to statistical mechanics with the aim to describe memory storage during development and in the adult. Networks with symmetric interactions have been shown to function as content-addressable memories, but the present approach differs from previous instructive models. Four biologically relevant aspects are treated--initial state before learning, synaptic sign changes, hierarchical categorization of stored patterns, and synaptic learning rule. Several of the hypotheses are tested numerically. Starting from the limit case of random connections (spin glass), selection is viewed as pruning of a complex tree of states generated with maximal parsimony of genetic information.

Learning↗

PubMedDate not supplied

Hierarchical neuronal modeling of cognitive functions: from synaptic transmission to the Tower of London.

Recent progress in the molecular biology of synaptic transmission, in particular of neurotransmitter receptors, offers novel information relevant to 'realistic' modeling of neural processes at the single cell and network level. Sophisticated computer analyses of 2D crystals by high resolution electron microscopy yield images of single neurotransmitter receptor molecules with tentative identifications of ligand binding sites and of conformational transitions. The dynamics of conformational changes can be accounted for by a 'multistate allosteric network' model. Allosteric receptors also possess the structural and functional properties required to serve as coincidence detectors between pre- and post-synaptic signals and, therefore, can be used as building blocks for a chemical Hebb synapse. These properties were introduced into networks of formal neurons capable of producing and detecting temporal sequences. In more elaborate models of pre-frontal cortex functions, allosteric receptors control the selection of transient 'pre-representations' and their stabilization by external or internal reward signals. We apply this scheme to Shallice's Tower of London test, and we show how a hierarchical neuronal architecture can implement executive or planning functions associated with frontal areas.

Allosteric Regulation↗

PubMedDate not supplied

The Wisconsin Card Sorting Test: theoretical analysis and modeling in a neuronal network.

Neuropsychologists commonly use the Wisconsin Card Sorting Test as a test of the integrity of frontal lobe functions. However, an account of its range of validity and of the neuronal mechanisms involved is lacking. We analyze the test at 3 different levels. First, the different versions of the test are described, and the results obtained with normal subjects and brain-lesioned patients are reviewed. Second, a computational analysis is used to reveal what algorithms may pass the test, and to predict their respective performances. At this stage, 3 cognitive components are isolated that may critically contribute to performance: the ability to change the current rule when negative reward occurs, the capacity to memorize previously tested rules in order to avoid testing them twice, and the possibility of rejecting some rules a priori by reasoning. Third, a model neuronal network embodying these 3 components is described. The coding units are clusters of neurons organized in layers, or assemblies. A sensorimotor loop enables the network to sort the input cards according to several criteria (color, form, etc.). A higher-level assembly of rule-coding clusters codes for the currently tested rule, which shifts when negative reward is received. Internal testing of the possible rules, analogous to a reasoning process, also occurs, by means of an endogenous auto-evaluation loop. When lesioned, the model reproduces the behavior of frontal lobe patients. Plausible biological or molecular implementations are presented for several of its components.

Frontal Lobe↗

PubMedDate not supplied

[Reading numbers in pure alexia: effects of the task and hemispheric specialization].

Selective conservation of the ability to read Arabic numbers in patients unable to read words or even letters is a classical characteristic of pure alexia described by Dejerine (1982). We report our work on the capacity of two patients with pure typical alexia to process numbers. Our main finding was that these patients could count pairs of Arabic numbers correctly when the reading task was simple (example 2 4-->"two four") or when the task involved comparing sizes (example 2 4-->"four is bigger than two"). Inversely, these patients often made mistakes when asked to perform arithmetic operations (example 2 4-->"two plus four equals six"). Using these two numbers, there was a similar dissociation between excellent performance on comparison tests and severe deficiency in reading out loud. We interpret these findings with the hypothesis that both of the hemispheres can identify Arabic numbers, but that the visual systems on the right and left play a different role during different tasks. In pure alexia, a lesion in the left identification system leads to selective deficiency in linguistic tasks such as reading numbers out loud, recognizing numbers with several figures or mental arithmetic. The right identification system in intact and is sufficient for comparison or reading isolated Arabic numbers.

Aged↗