Single motor unit mechanomyogram: a spike-triggered averaging approach

Cescon, Corrado and Nannucci, Lorenzo and Orizio, Claudio and Farina, Dario and Gazzoni, Marco (2002) Single motor unit mechanomyogram: a spike-triggered averaging approach. In: Proc. of the XIV Congress of the International Society of Electrophysiology and Kinesiology, 22-25 June 2002, Vienna, Austria.

Full text not available from this repository.


he purpose of this study was to extract single MMG MU response, using surface EMG detected potentials as triggers in an averaging process. The surface EMG signal was decomposed using wavelet transforms and a MUAP classification based on a neural network. The firings of the detected motor units were used as triggers to perform a sliding window averaging of the MMG signal. The deterministic component of the signal, extacted with the averaging process is the MMG response of single motor unit. The EMG signal was detected with a probe with 4 silver bars electrodes with 5 mm or 10 mm interelectrode distance depending on the muscle under study. The MMG signal was detected with a GS Sensors accelerometer model GSA3031, whose performances are comparable with the Analog Device model described in the contribution "Prototype of a hybrid probe for the joint detection of surface EMG and MMG signals". Low level contractions (10%-20% of the maximum voluntary contraction, MVC) were performed by the first dorsal interosseus, the biceps brachii and the tibialis anterior muscles. The result of the decomposition based on two double differential channels is the classification of the extracted potentials in different classes corresponding to the different motor units. Due to the superposition of the potentials, the complete decomposition is not possible; however the MMG response of single motor unit can be extracted in most cases. The presence of a deterministic MMG response was assessed with the asymptotic analysis of the spectral power of the averaged signals, increasing the number of firings on which the average is made. The minimum number of firings providing a good signal to noise ratio is about 30, the average length of the mechanical response varies between 20 and 40 ms depending on the examined muscle. The described technique has a potential relevance in studies of basic and applied physiology for the non invasive characterization of superficial muscles.

Actions (login required)

View Item View Item