Reproducibility of muscle-fiber conduction velocity estimates using multichannel surface EMG techniques.

The aim of this study was to assess the reproducibility of muscle-fiber conduction velocity (MFCV) estimates obtained from multichannel surface electromyographic (EMG) recordings. Surface EMG signals were collected with a matrix of 61 electrodes during isometric, submaximal (50% of the maximal voluntary contraction torque) contractions of the biceps brachii muscle. Conduction velocity was estimated using multichannel maximum likelihood techniques. Reproducibility of MFCV estimates was assessed varying the number of signals (two to seven) used for the estimate and the distance between detection points (5-30 mm). Intraclass correlation coefficient (ICC) of both initial MFCV values and their rates of change with fatigue increased when increasing number of signals and distance between detection points. ICC of initial MFCV was negative using two signals for MFCV estimate, and it increased to approximately 75% with six to seven signals. Thus, reproducibility of MFCV estimates may be improved significantly using advanced multichannel estimation methods with respect to classic two-channel techniques.