The measurement of neuromuscular junction jitter using a concentric needle electrode: a simulation study

Under the line-source model, ensembles of concentric-needle motor unit action potentials (MUAPs) were simulated. Neuromuscular junction jitter was simulated by delaying the initiation of muscle fibre action potentials (MFAPs) by a normal random variate. To transform the MFAP contributions to a MUAP into symmetrical, singly-peaked, positive functions, inverse-average-current filtering was applied to the MUAPs of the ensembles, producing composite weight functions. The jitter estimate, defined as the standard deviation of composite weight function interpeak intervals over an ensemble, was correlated to the standard deviation of the random variable used to delay the initiation of the MFAPs. Since normal jitter is of the order of 25 μsec, the base sampling interval of 40 μsec was enhanced to approximately 2 μsec using divided-difference interpolation. It was found that the measured jitter varied approximately linearly with the expected jitter. The correlation coefficients between the expected and measured jitter were 0.97, 0.96, and 0.94 for the three MUAP ensembles studied