Quantification of jiggle in real electromyographic signals

Abstract

Two parameters have been defined for quantifying jiggle: normalized consecutive amplitude differences (CAD) and the cross-correlational coefficient of consecutive discharges (CCC). In real recordings, artifacts from several sources may increase the variability of these parameters as they were originally defined. Two methodological modifications designed to overcome such a limitation are proposed: estimation of baseline fluctuation from segments of the recording free from nearby concurrent motor unit potentials (MUPs), and waveform alignment of consecutive discharges by correlation maximization (CM). The results obtained by the original and modified methods were compared for MUPs from normal subjects and patients with amyotrophic lateral sclerosis and chronic neurogenic diseases. With the modified method, CAD and CCC showed fewer extreme values and less scatter. The number of successfully aligned MUPs with the CM method was 18.8% higher (n = 394; Chi-square = 54.6; P < 0.001), including irregular and unstable MUPs. The proposed modifications improve our capability to quantify the jiggle of real signals and reduce the necessity of manual interventions although low-interference recordings and operator supervision are still required.