A Stochastic Model for Assessing Synchronization among Spike Trains in a Population of Motor Neurons

Synchronization among discharges in a population of motor neurons is of interest because of its potential to characterize physiological changes related to the neuromuscular system. Milner-Brown et al. (1973) developed a method to quantify synchronization in a population of motor neurons, in which synchronization is measured by averaging the spike-triggered surface electromyograms (EMG) waveforms. The surface EMG method opened a way to assess motor neuron synchrony in a large population of motor neurons instead of only a few, allowed investigators to track the same or similar groups of motor neurons longitudinally, and overcame the limit of examining only a few motor neurons using cross correlation. However, experimental results have suggested that the surface EMG method does not accurately and consistently detect motor neuron synchrony under some experimental conditions (Yue et al., 1995). This article reports our attempts to improve this method by establishing a new mathematical framework for the surface EMG procedure and to propose a general model based on this framework. The proposed model includes existing methods such as that of Hamm et al. (1985) as special cases. Based on the proposed model, we proposed a new synchronization index and performed computer simulation that indicated that the new index detects synchronization consistently with relatively high accuracy. Though based on the neuromuscular system, the proposed model should be extendable for detecting synchronization in other nervous systems.