Identification of muscle fasciculations from surface EMG: Comparison with ultrasound-based detection∗

The clinical standard for the identification of muscle fasciculations is needle electromyography. However, both surface electromyograms (sEMG) and ultrasound imaging (US) have been recently proposed as alternative and more sensitive approaches. The aims of this study were to: (i) compare the sensitivity to muscle fasciculations of sEMG and US, (ii) assess the rate of agreement (RoA) between the two approaches, and (iii) investigate how much sensitivity and RoA are affected by the selectivity of sEMG detection. Surface EMGs were collected concurrently with US images using an array of 32 electrodes spanning the whole, posterior aspect of the leg. Muscle fasciculations were identified from US videos and from monopolar and single differential sEMGs computed between electrodes spaced by 1, 2, and 3 cm. Results from five healthy subjects showed that US detected as many fasciculations as single differential EMGs, but always less than monopolar sEMGs. However, monopolar sEMGs exhibited a very poor spatial selectivity, likely responsible for the small RoA with US measures. The RoA was maximal for single differential recordings with 3cm inter-electrode distance, however, it was always smaller than 75% (median=30%). Although preliminary, these results suggest that sEMG and US are sensitive to different events in the muscle volume and that their integration may increase the detection sensitivity to muscle fasciculations.