Biomedical Engineering Reference
In-Depth Information
Statistical
methods
(n)
Decomposition
Features
Fourier, wavelet,
AR, cepstrum etc.
Segmentation/
identification
Physical waveform
measurements
No
Single
MUAP
Ye s
Rectification
Low-pass
filter
Bandpass filter
Bio amplifier
Muscle fiber
FIGURE 5.20
The major steps involved in the signal acquisition, preprocessing, and feature
extraction of EMG signals recorded from a muscle fiber.
signal. This preprocessing step reduces the amount of data to be processed
and potentially reduces the effect of noise. Noise in this context refers to
unwanted information from unimportant waveforms or background emissions
that superimpose on the required EMG signal.
EMG signals are first amplified, usually by a factor of 1000 since they
have a very small amplitude in the range of microvolts. It is preferable to
report on the characteristics of the amplification such as single, differential or
double-differential amplifiers, input impedance, common mode rejection ratio
(CMRR), and SNR because the following processing stages, such as template
matching, may require this information.
The next signal-processing method is filtering and MUAP segmentation.
In some experiments, for example, the EMG signal is high-pass filtered at
250 Hz to remove background noise (Pattichis et al., 1995). The beginning
Search WWH ::
Custom Search