Indeterminate organization of the visual system

Science

Volumn 271, Issue 5250, 1996, Pages 776-777

  Hilgetag, Claus C ^a   O'Neill, Mark A ^a   Young, Malcolm P ^a  

^a NEWCASTLE UNIVERSITY   (United Kingdom)

Author keywords

[No Author keywords available]

Indexed keywords

ALGORITHM; HUMAN; NERVE CELL NETWORK; NEUROANATOMY; PRIORITY JOURNAL; SHORT SURVEY; SYNAPSE; VISUAL CORTEX; VISUAL SYSTEM;

ALGORITHMS; ANIMALS; BRAIN MAPPING; MACACA; NERVE NET; NEURAL NETWORKS (COMPUTER); VISUAL CORTEX; VISUAL PATHWAYS;

EID: 0030027057     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.271.5250.776     Document Type: Short Survey

References (11)

1. K S Rockland and D N Pandya, Brain Res. 179, 3 (1979), J H R Maunsell and D C Van Essen, Annu Rev. Neurosci 10, 363 (1983), T A Coogan and A Burkhalter, J Neurosci 13, 3749 (1993)
2. K S Rockland and D N Pandya, Brain Res. 179, 3 (1979), J H R Maunsell and D C Van Essen, Annu Rev. Neurosci 10, 363 (1983), T A Coogan and A Burkhalter, J Neurosci 13, 3749 (1993)
3. K S Rockland and D N Pandya, Brain Res. 179, 3 (1979), J H R Maunsell and D C Van Essen, Annu Rev. Neurosci 10, 363 (1983), T A Coogan and A Burkhalter, J Neurosci 13, 3749 (1993)
4. D J Felleman and D C. Van Essen, Cereb. Cortex 1, 1 (1991)
5. Nodes in a conventional neural network have fixed connections with variable weights. Computation consists of numerical operation on an input vector to generate an output vector. The data in our case are not numbers but relational constraints We generalized the properties of the simulated network so that each node could be associated with a variable position in the network and could have a relation with other nodes, or none at all, as defined by the input constraints. This organization allowed the structure of the simulated network itself to stand for the network in the real brain, rather than the representational states being patterns of numbers distributed over the nodes These features might extend the processor's use to other relational problems, such as macromolecular conformation.
6. W Metropolis et al , J Chem. Phys 21, 1087 (1953), P. J M Van Laarhoven and E H L Aarts, Simulated Annealing: Theory and Applications (Kluwer, Dordrecht, 1987)
7. W Metropolis et al , J Chem. Phys 21, 1087 (1953), P. J M Van Laarhoven and E H L Aarts, Simulated Annealing: Theory and Applications (Kluwer, Dordrecht, 1987)
8. MDP and MIP were excluded from the analysis.
9. The relatively low number of violations in (2) suggests that this hierarchy may be in the top 100 million hierarchies.
10. D Boussaoud, L G Ungerleider, R. Desimone, J Comp Neurol 296, 462 (1990)
11. Supported by the Medical Research Council, the McDonnell-Pew Foundation, Newcastle University and the Royal Society