Neural and Mathematical Modleling
of Neuroimaging Data

Neuroimaging has become an invaluable tool for cognitive science, clarifying brain regions that are involved in specific cognitive function. However, there is a danger in blindly assigning "labels" known in current cognitive science to citoarchitecturally classified brain areas. Indication of problem in such approach is seen in the fact that the same brain regions are repeatedly implicatied in different cognitive functions. In fact, we don't know whether we should best label certain area of a brain as "semantic area" or "array of shift registers". To reach an understanding of brain mechanisms underlying cognitive functions, we need an mechanistic model that explain how different brain regions cooperate to carry out cognitive functions.

That is why neural modeling may help brain imaging data analysis. Neural network modeling seeks to provide explanatory theories of how the brain works. On the otehr hand, neural network modeling is also entering another era. Although connectionist models have been useful at a metaphorical level in explaining cognitive functions, researchers are aware that there are much work needed to move forward and neural network models will become models of the biological neural structures. Recently, ion channel level neuron simulators have become widely available and efforts to link this level to large scale neural networks are underway. There have also been an active research linking neural activities to electrophysiological and hemodynamic signals.

This webpage is a directory of links and reference related to modeling of neuroimaging data. The collection heavily reflects research interest of the site manager . Please email him if you know cool links and papers that should be added!