**1. Introduction**

Computational Intelligence in Electromyography Analysis – 390 A Perspective on Current Applications and Future Challenges

Zollo L., Roccella S., Guglielmelli E., Carrozza M. C. & Dario P. (2007). Biomechatronic design and control of an anthropomorphic artificial hand for prosthetic and robotic

applications, *IEEE/ASME Transactions on Mechatronics*, Vol. 12, No. 4, pp. 418-429.

Surface electromyography (sEMG) is a generic term for a method of recording electrical muscle activity. Numerous applications for this method have been developed in clinical practice, such as diagnosing neuromuscular diseases, analyzing and determining abnormalities or disorders and muscular rehabilitation (biofeedback) [3, 12, 27, 28].

sEMG is mainly used in the fields of physiotherapy, dentistry, physical education and biomechanics [12].

The duration of sEMG activity corresponds to the duration of muscle activation. The amplitude is the level of signal activity and varies with the amount of electrical activity detected in the muscle; it provides information about intensity of muscle activation. The observed sEMG frequency is due to a wide range of factors: muscle composition, characteristics of the action potential of the active muscles fibers, the intramuscular coordination process and electrode properties [22, 23, 28].

sEMG signals are also affected by the anatomical and physiological properties of the muscles, neuromuscular control of the peripheral nervous system and the instrumentation used to collect the signal.

The electronic EMG device amplifies, isolates and filters the electrical signal of muscles that occurs during muscle contraction. This signal must undergo conditioning to be captured [12].

A differential amplifier is, ideally, insensitive to noise and amplifies only the EMG signal, although in practice this is not the case. This situation occurs, first of all, because the noise that reaches the electrodes (inputs) doesn't necessarily have the same magnitude. Moreover, due to technological limitations, differential amplifiers cannot perfectly separate two-signal input.

The measurement that indicates the success of this separation is the common mode rejection ratio (CMRR), which is usually expressed in decibels (dB). The CMRR value of the differential amplifiers used in sEMG is on the order of 80 to 100 dB [3, 22, 24].

Application of Surface Electromyography in the Dynamics of Human Movement 393

monitoring the electrical signals from muscle contractions. A major methodological problem for EMG is the frequent presence of artifacts or noise. Artifacts or noise are defined as information whose origin is distinct from the neuroelectrical muscle activity signal. Some examples of this include interference, heart rate, poor contact between the electrode and the

The presence of artifacts is difficult to avoid with this type of signal acquisition, since in order to amplify the signal, which is received in microvolts (μV), unwanted signals are also amplified and can compromise interpretation of the EMG signal. Thus, the signal-to-noise ratio has been a problem, and numerous studies have been undertaken to resolve EMG signal interpretation problems. After several attempts, a solution was found in the development of the differential amplifier [3] (ACIERNO, BARATTA & SOLOMONOW,

The signal amplifier is an electronic device that filters, amplifies and records bands of

The initial problem with the amplifiers was that signal acquisition was dependent on the electrical resistance of the skin. Thus, in many studies skin resistance and temperature were initially monitored when the test was performed, conditions that made it difficult or

Over time, corrections have been made to this system so that the amplifiers currently have high input impedance and attenuate noise levels, which allows the reproduction of

A main feature of this new generation of amplifiers is that they can amplify a particular type of biological signal independent of skin resistance [28]. The evolution of cables and connectors must also be considered in the development process of EMG acquisition equipment, since the type of conductive material and insulation system help minimize

The main purpose of these developments is to help investigate and analyze human movement. The field of biomechanics is a practical example of the use of technological

Biomechanics can be defined generally as the study of the mechanics of living beings, or more specifically, the science that examines forces acting upon and within a structure and the biological effects produced by these forces [17]. Given the complex approach involved in biomechanics and human movement analysis [17], it is important to discuss the concepts, criteria and methods involved, focusing on the use of EMG for reliable interpretations.

EMG can be defined as the study of muscle function by analyzing the electrical signal generated during muscle contraction. Studying muscle function by means of EMG can be carried out under both normal and pathological conditions [12]. EMG has been used in important studies on muscle activity that have both qualitatively and quantitatively addressed the function of human movement. New information about muscle activity has

impossible to reproduce and some EMG experiments [1].

experiments without interference with the results.

resources to interpret human movement [28].

skin, etc.[12].

1995).

signals.

noise.

The sEMG equipment should be calibrated before recording signals. Calibration is important for fidelity, accuracy and reliability when reading the signal. The amplification factor is critical during the calibration process, since it is the ratio between the input voltage and that which comes out of the amplifier. The gain is selected according to the requirements of the type of experiment, the studied muscles, the type of electrodes involved and the planned use of the amplified signal. Whereas an sEMG signal has a maximum voluntary contraction amplitude not exceeding 5 millivolts (mV) peak-to-peak, the gain should be adjusted to 500-1000x [2,3,5].

During the mathematical processing of the sEMG signal, filters can be used to remove components that don't belong to the signal or components that are irrelevant for a given analysis.

The useful information in the sEMG signal is located in a particular frequency band (20-500 Hz), and is reduced by a filtering effect from the tissue located between the muscle fibers and the active sensing surface. The filter band corresponds to the frequency between the low- and high-cut filter frequencies [28].

Time-based signal processing can be carried out using a set of processing procedures intended to characterize the signal's curve and measure signal strength during the contraction. Signal processing applications in the time domain are widely used in areas such as neuromuscular coordination, motor control, the relationship between EMG and strength and muscular coordination in the dynamics of human movement.

This chapter will report, therefore, on the importance of sEMG with respect the dynamics of human movement [27].
