Sensor Interface for Electromyogram with Wavelet Filter

Abstract

This paper presents electromyogram(EMG) sensing interface with wavelet process. Wavelet process is effective to analyze EMG signal because it provides signal information in frequency domain. The power of EMG signal is concentrated in 50~150Hz range. By using wavelet transform filters, the EMG signal is decomposed in several ranges of frequency. In this process, immunity to motion artifacts is enhanced and analyzing an EMG with various frequency ranges gives more information about specific movement. the amount of data is decreased because down sampling is used in wavelet transform process. Thus, it improves communication efficiency. The proposed interface is designed in 0.18μm CMOS process.