Comparison of three algorithms for multi‐motor unit detection and waveform marking

Abstract

Quantitative EMG (QEMG) techniques include automated motor unit action potential (MUAP) detection and marking of clinically useful waveform metrics. Different computer algorithms are available on modern EMG machines to perform these operations rapidly. However, the efficiency and accuracy of available algorithms are rarely directly compared. We have assessed three commercially available algorithms using both synthesized and biologic interference patterns and found differences among algorithms, some of which are clinically significant. Our results point out the importance of assessing for duplicate MUAPs (same waveform detected as two separate waveforms) and accuracy of markings used to determine MUAP metrics.

Optimizing acquisition time in quantitative electromyography

“…1). This and a previous study2 have shown that an average of 6 distinct MUAPs per contraction is a practical upper limit; this number of MUAPs can typically be achieved with acquisition times of 10 seconds. Shorter data acquisition epochs may lead to inadequate numbers of individual MUAP firings due to the occasional superimposition of different MUAPs.…”

“…Automated QEMG algorithms rely on the ability to identify similar MUAPs in a discharge train to create a composite MUAP from which noise is essentially removed and which allows for more accurate marking of metrics. Data acquisition time is an important variable in order to collect a sufficient number of different MUAPs and to average them 2. Commercially available QEMG algorithms have differing requirements for the number of similar MUAPs needed to compile the composite MUAP used for averaging.…”

Averaging methods for extracting representative waveforms from motor unit action potential trains

“…Ensemble (simple) averaging (EA) (Sörmo and Lagunas, 2005) is the most common approach applied for extracting representative waveforms from trains of repetitive potentials and has been used with different sorts of signals: evoked potentials in EEG (Vedel-Larsen et al, 2010), cardiac cycles in electrocardiography (Alperin and Sadeh, 1986) or MUAPs in EMG (Brownell et al, 2006). The EA technique is simple and intuitive and gives satisfactory results at low and uniform levels of noise.…”

EMGTools, an Adaptive and Versatile Tool for Detailed EMG Analysis

“…make it difficult to generalize or put restrictions on the raw signal, the MUAPs, or the FPs, especially in recordings from pathological muscles. We therefore prefer to use real rather than synthesized signals [27]- [30] during development, optimization, and evaluation of EMGTools.…”