Bondarenko Y. Statistical modeling of dendrites of neurons

In dissertation the probability model and the simulation algorithm of dendrite neuron are offered. New concepts and characteristics of dendrite are entered for the description of dendritic structure and for the construction of its model. The algorithm of dendrite's modeling consists of the algorithm of modeling of segment's link, the algorithm of modeling of segment without subtrees, the algorithm of modeling of segment with subtrees, the algorithm of modeling subtree. The influence of the basic numerical characteristics of probability model on the dendritic form and on the values of the emergent characteristics of dendrite is investigated. The work of algorithm and the adequacy of probability model of dendritic tree are verified on the example of Purkinje cells from guinea pig cerebellar cortex. The adequacy of the description of real dendrite by model is established by check of statistical hypotheses about concurrence of distributions and parameters of distributions of the basic and emergent numerical characteristics of model and real dendrite.