Signal Classification and Electromyography (EMG) Instrumentation Design as Basic Electronic Control System

Abstract

Human body naturally contains complex electrical signals ranging from the mechanism of the brain, heart, and muscles. Electromyography (EMG) is the development of biomedical engineering devices used to record electrical signal activity produced by muscles. The signal produced by the muscle has a very small amplitude, therefore, an amplifier circuit is required to read the signal using microcontroller. In this study, EMG signal amplification was carried out periodically to reduce errors that might occur if only charged to one signal amplifier. Signal reading was performed by conducting 5 different basic hand movements. Muscle signal activity during relaxed conditions had an amplitude of 0.05 V. Moreover, it conveyed an amplitude of 0.27 V up to 2.62 V with a signal gain of 10.350 when the muscle did different movements. The K-NN algorithm was used to classify EMG signals and determined the hand motion output used as a basic electronic control system. Results showed success rate of 88%. Failures when carrying out basic electronic control was due to the signal classification processes that had not met the parameters and the signal amplitude that was almost similar to another signal.

Classification of Electromyography Signal with Machine Learning

“…The sensor has a micro-sensing metal that can be combined with ECG measurement and is suitable for electromyography signals with low-quality reception that does not require multiple gesture recognition. There is also a signal wire connecting the electrode [9] to the sensor as shown in Figure 2. ESP8266 microcontroller.…”