On the experimental results of a noninvasive estimation technique of muscle conduction velocity distribution

Abstract

-This paper is intended to present an assessment of the experimental results of a previously proposed muscle conduction velocity distribution (CVD) estimator. The performance of the proposed technique was seriously deteriorated when applied to experimental data. The causes for this decline were evaluated by introducing real-world errors in the model parameters and looking at how sensitive the estimator is to these. The simulation results show the high sensitivity of the estimator to parameter errors. The similitude found between the simulation results and the CVD estimates obtained on experimental data helped confirm this observation as well. Given the sensitivity observed the proposed technique is not practical for muscle dimensions such as those found in the biceps brachii. Keywords -Conduction, propagation, velocity, CVD, EMG, noninvasive, estimation. I. INTRODUCTIONNoninvasive techniques for the estimation of conduction velocity distribution (CVD) on human muscles can lead to new possibilities for clinical assessment of muscular pathologies. Efforts are currently directed to the development of such tools [1] [2]. However due to the constraints imposed by the low frequency content of the electromyographic (EMG) signal together with physiological issues related to muscle geometry this task is a challenging one. This is true particularly when the aim is to characterize a representative part of the muscle motor unit population through a CVD estimate. II. METHODThe technique used by the authors is a deconvolution approach based on volume conductor modeling of the surface EMG interference pattern [2]. It makes use of two correlation functions computed from the surface EMG data recorded at the skin surface with a bipolar configuration. A volume conductor model of the signal that depends on different parameters is built into the estimator. Some of these parameters such as interelectrode spacing, which is a rather accurate quantity, are determined by the recording configuration. However, other parameters such as 1. Location of the recording electrodes with respect to the innervation zone, and 2. Muscle geometry and thickness of the fat layer under each recording channel, are not known with certainty.The setup shown in Figure 1 was used for experiments and simulations. Two bipolar channels spaced by a distance dch were aligned in the direction of the muscle fibers. As part of the tests implemented through simulations, the distance from the recording electrode arrangement to the middle of the innervation zone was underestimated by 5, 10 and 20mm during the CVD estimation. The power of the signals recorded from channels X & Y can also differ from that expected according to the assumption made on model parameters mentioned in point 2. These differences in channel powers convey differences in the relative amplitudes of the MES auto and cross correlation functions R XX and R XY , respectively. The effect that this has on the estimator performance was evaluated by upsetting the auto-to-cross amplitude ratio of the ...