f Review Methods for estimating muscle fibre conduction velocity from surface electromyographic signals \ D. Farina R. Merletti Dipartimento di Elettronica, Laboratorio di Ingegneria del Sistema Neuromuscolare (LISiN), Politecnico di Torino, Torino, Italy Abstract--The review focuses on the methods currently available for estimating muscle fibre conduction velocity (CV) from surface e/ectromyographic (EMG) signals. The basic concepts behind the issue of estimating CV from EMG signals are discussed. As the action potentials detected at the skin surface along the muscle fibres are, in practice, not equal in shape, the estimation of the delay of propagation (and thus of CV) is not a trivial task. Indeed, a strictly unique definition of delay does not apply in these cases. Methods for estimating CV can thus be seen as corre- sponding to specific definitions of the delay of propagation between signals of unequal shape. The most commonly used methods for CV estimation are then reviewed. Together with classic methods, recent approaches are presented. The techniques are described with common notations to underline their relationships and to highlight when an approach is a generalisation of a previous one or when it is based on new concepts. The review identifies the difficulties of CV estimation and underlines the issues that should be considered by the investigator when selecting a particular method and detection system for assessing muscle fibre CV. The many open issues in CV estimation are also presented. Keywords--Electromyography, Delay estimators, Spatial filters, End-of-fibre compo- nents, Conduction velocity Med. Biol. Eng. Comput., 2004, 42, 432-445 J 1 Introduction MUSCLE FIBRE conduction velocity (CV) is an important physiological parameter because it reflects the membrane muscle fibre properties and thus the modifications of the peripheral properties of the neuromuscular system as a conse- quence of pathology (VAN DER HOEVEN et al., 1993, 1994; LINSSEN et al., 1991; ZWARTS et al., 2000), fatigue (BIGLAND- RITCHIE et al., 1981 ; ARENDT-NIELSEN et al., 1989; MERLETTI et al., 1990), pain (FARINA et al., 2004b) or exercise (SADOYAMA et al., 1988; RONGEN et al., 2002). CV can be estimated from either intramuscular or surface electromyo- graphic (EMG) signals (ARENDT-NIELSEN and ZWARTS, 1989). in this work we will focus on the estimation methods applied to surface EMG signals. Estimation of CV from surface EMG recordings is a complex task. it requires signal detection with advanced systems and experimental procedures and the analysis of signals corrupted by various sources of noise. The basic requirement for the estima- tion of CV is the detection of action potentials along their propagation from the innervation zone(s) to the tendon Correspondence should be addressed to Dr Dario Farina; emaih dario.farina@polito.it Paper received 17 November 2003 and in final form 26 February 2004 MBEC online number: 20043893 © IFMBE: 2004 432 regions, ideally, the surface-detected signals should be unchanged in shape along the fibre length, but this condition is not met in practice. There are many reasons for changes in the shape of the surface detected signals along the muscle fibre direction. First, the detection system may not be perfectly aligned with respect to the muscle fibre orientation. Secondly, the potentials do not purely propagate along a line but rather are generated and extinguished at the end-plates and tendons. Moreover, for each muscle fibre, there are two potentials propagating in opposite directions to the two tendon regions. Thirdly, each motor unit (MU) has a specific CV of propagation, and thus the interference surface EMG signal is comprised of potentials travelling at different velocities. The resultant signal cannot be characterised as purely propagating when detected along the muscle fibre direction (this is an issue when average CV, rather than single MU CV, is estimated). Fourthly, the tissues separating the muscle fibres and the recording electrodes are inhomogeneous along the direction of propagation, thus affecting the surface signal shape during the propagation. Fifthly, the signals always contain noise and are sampled at discrete instants of time rather than acquired as analogue functions, in some conditions, the factors affecting the shape of the action potentials can be critical, which can hinder the possibility of estimating muscle fibre CV from surface EMG signals. This is, for example, the case with multi-pinnate muscles, muscles presenting largely scattered innervation zone locations, or muscles covered by thick subcu- taneous layers (MASUDA and SADOYAMA, 1987). Medical & Biological Engineering & Computing 2004, Vol. 42