Biometrics and Healthcare System Using EMG and ECG Signals

Abstract

In this paper, we study biometrics and healthcare system using electromyogram (EMG) and electrocardiogram (ECG) signals. Most people use smart devices for convenient life and they should be authenticated to access their devices. In this course, people should give their bio-signals to the devices inevitably. Here we use EMG signal for biometrics and ECG signal for heart disease detection. The EMG and ECG signals are a bio-signal that measures the potential difference generated by muscles and heart through the skin of the body surface, respectively. Using the EMG signal for biometrics improves the security of visual exposure. However, the EMG signals show various shapes by muscular strength and which muscles are used. For designing robust biometrics using EMG signal, a set of useful features is extracted and the features enter long short-term memory for training the network. For heart disease detection using ECG signals the signals are transformed into an image and the image is used to train a convolutional neural network. Publicly opened datasets of EMG and ECG signals are used for the experiment and the performance is compared by some parameter changes.