(Prof.V.G.Yakhno, Dr.I.V.Nuidel) In development of technical versions of neurocomputer systems it is important to take into account and effectively use appropriateness of biological prototypes. A computer model of input sensor signal processing in the interconnected homogeneous neuron-like systems is developed. It is considered on the basis of the well-known architecture of a sensor signal passing in neural ensembles of animals. In well-known architecture of a sensor signal passing in neural ensembles of animals interconnected homogeneous neuron-like systems such as “cortex-thalamus-reticular thalamic neurons” chains is considered. (Fig1.). The model allows for analysis of qualitatively different functioning regimes which are solutions of the model.

yakhno-r3-rus

Fig. 1. (a) External sensor signal interactions and pathways to the cortex, thalamus and neurons of the reticular-thalamic formation. (b) Schematic of transmission of external signal to cortex and interaction between the cortex, thalamus, and neurons of the reticular-thalamic formation which is used for model calculations. In normal regime brain processes two-dimensional information flows presented as patterns activity of neuron ensembles. Consideration of the dynamics of two-dimensional patterns activity, search qualitative peculiarities of sensor information processing in normal regimes, and breakdown condition of normal functioning regimes are very important for simulation and interpretation of different functioning system regimes. For example, it is known “dynamic diseases” exist. And normal living tissue can give a wide variety of abnormal dynamic regimes such as self-activation regimes. These regimes belong to epileptiform seizure activity in the brain. More than forty variety of seizures in brain are recorded experimentally. Developed model for interacting neural ensembles has demonstrated possibilities for understanding of mechanisms of qualitatively different functioning regimes some of which can be interpreted as normal operation of extracting information features and others as pathological regimes (similar to spike wave discharges) inhibiting normal processing of sensory data. Adequacy of this model is confirmed by results of qualitative agreement with experimental data.

Topics of term papers

  1. The model of homogeneous distributed neuron-like system as a model of receptive fields of visual system for stimulus feature isolation.
  2. Research of spatial temporal regimes of sensor information transformation in three-component neuron-like system of interconnected “cortex-thalamus-reticular thalamic neurons” chains.
  3. Modeling of processes of sensor information coding-decoding on the base model of homogeneous distributed neuron-like system.