ABSTRACT
In this chapter, the authors discuss a digital signal processor (DSP)-based system that integrates a signal preprocessing module, a digital filter module, and a pattern recognition module. The DSP integrates the signal preprocessing module, the digital filter module, and pattern recognition module into the neural controller. The on-line DSP neural controller can provide an 87.5% correct rate for the discrimination of eight hand motions. In order to avoid the distortion of converting from the analog electromyographic (EMG) signal to digital signal, the EMG signal is further amplified about twenty times using an operational amplifier. The EMG signal is used to discriminate eight hand motions: power grasp, hook grasp, wrist flexion, lateral pinch, flattened hand, centralized grip, three-jaw chuck, and cylindrical grasp. In order to choose meaningful EMG signals for eight kinds of prehensile postures, the location of electrodes is important. The pattern recognition is implemented in the DSP-based prosthetic controller system.
