The long term objective of the present research is to provide accurate mathematical, computational, and theoretical models of the functional architecture of the neo-cortex.
The specific aims are to build the basic computational and theoretical tools and models required for a quantitative understanding of functional architecture in sensory cortex. Two aspects of this work are described in the present proposal: 1.)A prototype system for reconstructing cortex in 2-D and 3-D, from serial sections, has been constructed. A second generation software system for computer aided neuroanatomy will be constructed in which human interactive requirements and computer intensive stages of prototype system will be minimized. This work will be publicly distributed, along with the earlier prototype system. Serial section data sets, for the purpose of testing, calibrating, and learning to use the software tools, will also be distributed. 2.) Pilot data is shown which indicates that an accurate and concise model of the primary visual cortical map and ocular dominance column system can be constructed. It is proposed to develop these pilot studies into a full model and simulation of topographic and columnar architecture. This work represents a computational approach to the description of anatomical patterns. By standardizing and distributing data sets, software, and descriptive techniques, animal usage can be reduced. It provides fundamental methods and algorithms for reconstructing both the biological data and the neuronal patterns which are contained within that data. A careful measurement and characterization of these patterns is of critical importance to full understanding of the relationship between brain and behavior.
| Tal, D; Schwartz, E L (1997) Computing with the leaky integrate-and-fire neuron: logarithmic computation and multiplication. Neural Comput 9:305-18 |
