Various research activities have been conducted in the field of electromyography and few of its applications in specific areas are presented here:
2.1 An EMG-Controlled Graphic Interface Considering Wearability
Here EMG signal is used to control a graphical interface by classifying the wrist motion using fuzzy min-max neural network.
“Human Computer Interaction (HCI) technology using a bioelectric signal such an electromyogram (EMG), an electroencephalogram (EEG) and an electrooculogram (EOG) is considered an alternative to conventional input devices such as a keyboard or a mouse. Among these bioelectric signals, an EMG can be used as a control source for an intuitive and natural HCI because EMGs represent electrical activity of muscles.”
2.2 Bioelectric Control of a 757 Class High Fidelity Aircraft Simulation
It demonstrates bioelectric flight control of 757 class simulation aircraft landing at San Francisco International Airport. The physical instrumentality of a pilot control stick is not used. A pilot closes a fist in empty air and performs control movements which are captured by a dry electrode array on the arm, analyzed and routed through a flight director permitting full pilot outer loop control of the simulation.
“The basic block shows the steps used to record, extract and sense flight control gestures. The Bioelectric signals passed through three steps to be interpreted as gestures: pre- processing, feature vector formation, and gesture recognition. The pre-processing stages consist of electrode placement and filtering. The remaining two steps, forming feature vectors and recognizing gestures, have many potential implementations. Typical features used for signals includes moving averages of absolute values, neural network weights, Auto-Regression (AR) and mel-cepstrum coefficients, principal components, weighted factors, measures of slope and acceleration of transform coefficients, Short-Time Fourier Transforms (STFT), and wavelet decompositions. Pattern recognition models that might be used to perform pattern recognition include simple threshold indicators, neural networks, linear separators, or HMMs.”
2.3 Using Singular Eigen values of Wavelet Coefficient as the input of SVM to recognize motion Patterns of the Hand
“Considering the non-steady character of electromyography signal, wavelet transform is employed to analyze electromyography on the basis of acquired signals that have been preprocessed earlier, consequently singular value decomposition of a wavelet coefficient matrix is adopted to extract features of surface electromyography and the Directed Acyclic Graph Support Vector Machine algorithm is utilized to implement the multi-motion pattern recognition of surface electromyography.”
2.4 Wavelet Analysis of Surface Electromyography Signals
“Fast Fourier Transform (FFT) is one of the most common methods for analyzing the signal whether it is filtered or not. Another DSP technique is referred to as Wavelet analysis, a method that is gaining more use in analyzing SEMG signals. Both DWT and WPT use analytical wavelets called “mother wavelet,” which comes in different sets or “families.” Wavelet analysis has the advantage over FFT as it provides the frequency contents of the signal over the time period that is being analyzed. SEMG signals were collected from a muscle under sustained contractions for 4 seconds with different loads. The raw signals were analyzed using FFT, DWT and WPT in Lab VIEW® using its Signal Processing Toolset. Using Wavelet analysis the SEMG signal was decomposed into its frequency content form and then was reconstructed.”