The Impact of Using Digital Filter and Analog Filter on Surface Electromyography Signal

Abstract

Many accident cases result in humans having to going a surgery to save them, then performing muscle therapy to help the patient’s recovery after going through the post-surgery. The purpose of this therapy is the patient’s body to its normal state. Exoskeleton is an additonal clothing-like tool that aims to both protect and increase the wearer's abilities. Meanwhile electromyography (EMG) is a technique to evaluate and record the electrical activity produced by skeletal muscles. The purpose of this study was to analyze the differences in using of analog and digital filters on EMG, as well as the effect on the exoskeleton simulation. The method used in the main design consists of the myoware module, notch circuit, low pass filter, arduino uno, DAC module, teraterm software, and matlab. The intercepted signal was taken from the biceps using a disposable electrode (AG/AGCL.). The EMG signal tapped by the myoware module then is continued to another circuit, then was recorded on the Teraterm software, and analyzed in MATLAB. The voltage value on the analog filter is 1.541 Volt during relaxation and 2.086 Volt during contraction, while the digital filter that has passed through the DAC has a value of 41.8 mVolt during relaxation and 269.1 mVolt during contraction. The results of this study obtained that digital and analog filter values ​​have an average difference of 5 to 30. The conclusion of this research is that the tool can detect changes in the use of analog and digital filters. Therefore, in the future research, development can be made to compare other types of digital filters along with replacement to wireless systems. The benefit or purpose of this research is as a simulation of exoskeleton skeletal motion and to see the difference between the use of digital and analog filters.

Implementation of a neural network of low computational cost for its application in arm prostheses

“…For the implementation of a pattern recognition ANN, a training phase is required where the data to be learned and the category to which it must be assigned are presented. Then, for the first stage, the EMG signals from the biceps are sampled at three voltage thresholds (Cote-Allard et al, 2019;Hagengruber et al, 2022;Setioningsih, 2021). Figure 3 shows the thresholds distinguished by vertical dotted lines, on the left "low voltage", in the middle "medium voltage" and on the right "high voltage".…”