SYNCHRONY DETECTION IN NEURAL ASSEMBLIES

The identification of synchronously active neural assemblies in simult aneous recordings of neuron activities is an important research issue and a difficult algorithmic problem. A gravitational analysis method h as been developed to detect and identify groups of neurons that tend t o generate action potentials in near-synchrony from among a larger pop ulation of simultaneously recorded units. In this paper, an improved a lgorithm is used for the gravitational clustering method and its perfo rmance is characterized. Whereas the original algorithm ran in n(3) ti me (n = the number of neurons), the new algorithm runs in n(2) time. N eurons are represented as particles in n-space that 'gravitate' toward s one another whenever near-synchronous electrical activity occurs. En sembles of neurons that tend to fire together then become clustered to gether. The gravitational technique not only identifies the synchronou s groups present but also shows graphically the changing activity patt erns and changing synchronies.