Analysis of striatal dynamics: the existence of two modes of behaviour

The qualitative dynamical behaviour of a neural model based on the mammalian neostriatum was analyzed. The neostriatum was modelled as a mutually inhibitory network of physiological neurones, which was driven by excitatory afferents from the cerebral cortex. The analysis defined the conditions under which the system would enter into one of two dynamic modes, competition or co-activation, in terms of the parameters defining receptor-operated and voltage-sensitive channels in the neuronal membrane. We have previously argued that the mode of co-activation in the neostriatum may correspond to the state of muscular rigidity which occurs as a symptom of Parkinson's disease. The present work extends a preliminary analysis of a two-neurone system to a system of arbitrary size. An explicit prediction is made of the conditions under which a transition from co-activation to competition will occur, which is testable experimentally. The wavelength of a non-uniform activity pattern produced by small departures from uniform afferent drive is determined for one- and two-dimensional arrays of neurones. Two mild assumptions about the connectivity of the network were used to simplify the analysis, namely that the network was symmetric and homogeneous. The implications of departures from these assumptions for understanding the disordered movement seen in Huntington's disease are also considered.