Detecting Multi‐Neuronal Discharge in Human Muscle Sympathetic Nerve Activity

Detection of multiple neurons from human muscle sympathetic nerve activity (MSNA) is difficult to achieve in raw data. Based on wavelet denoised data, a spike detecting algorithm is proposed on which burst specific features of multiple single neurons are analyzed including spike count and distributions of spike amplitude and inter‐spike interval. Raw MSNA data (microneurography) were recorded from a single subject during baseline and during a maximal end‐inspiratory breath‐hold. Data sets each containing 10 bursts were selected from the baseline and each reflex period. The total spike count over 10 bursts at baseline (136 spikes) increased to 240 spikes during the inspiration reflex. During the inspiration phase the range of spikes detected under a burst was 16–36 which was higher than at baseline (6–25). Reflex mean spike amplitude (0.36±0.01 v) did not vary with baseline (0.38±0.01v), but inter‐spike interval (19.2±0.3 ms) was different from baseline (26.8. ±0.3 ms). Distributions of spike amplitude and inter‐spike interval in single bursts were similar if the number of spikes were similar and above 20. Moreover, the distributions over separate 10‐burst segments appear to be stationary. The number and frequency of spikes, not their amplitude, determined the integrated neurogram burst amplitude. The preliminary results indicate that the proposed spike detecting algorithm is useful for identification of multiple action potentials within each burst and for characterizing the burst specific features in human multiunit MSNA records. Supported by NSERC‐CHRP