Fast-slow phase separation of near infrared spectroscopy to study oxigenation v/s sEMG changes

Abstract

The possibility of studying non-invasively local muscle oxidative metabolism during exercise has been recently enhanced thanks to the use of Near-InfraRed Spectroscopy (NIRS). Moreover, the myoelectric manifestations of muscle fatigue occurring during sustained isometric contractions have been extensively studied with quantitative surface electromyography (sEMG) and are described by means of some sEMG global parameters extracted in time- and/or frequency-domain. With the aim of analyzing together both NIRS and sEMG data recorded on Biceps Brachii, an experimental protocol has been applied to seven subjects. During the experimental session, each subject had to perform two different kind of physical trials, a constant force and a cyclically varying isometric contraction. While examining the whole set of data with the aim of investigating the relationship between modifications of sEMG and underlying metabolic status, we faced up with the problem of objectively separating the two main phases of NIRS data. Results clearly indicated the presence of an initial fast phase of muscle desaturation followed by a slow phase, regardless of the kind of exercise. This behavior was paralleled by an analogous rate of change of sEMG parameters, thus suggesting a strong link between the two phenomena. As an objective criterion for separating the fast and the following slow phase an ad-hoc algorithm has been implemented. NIRS & sEMG data have been analysed by considering only those data pertaining rigorously to one of the two phases, whereas data collected during initial, transiction and final phases were discarded. In this paper we present some details on the algorithm for automatic separation of the two phases, with the most important results on statistical significance of the relationship between parameters extracted from sEMG in the fast phase v/s muscle oxigenation.