Affiliation:
1. GAZI UNIVERSITY, INSTITUTE OF SCIENCE
2. GAZİ ÜNİVERSİTESİ
Abstract
In this study, a wearable electromyogram (EMG) system on the forearm was designed to analyze finger movements for use in human-machine interfaces. The designed system measures the EMG signals without restricting the user's movements, analyzes these measurements through the software embedded in the system, and transmits the generated response to the output units to be controlled in real-time with wireless communication techniques. In the study, a three-channel EMG amplifier was designed and a system in which the NodeMCU V3 development board could be integrated was realized.With the system, the features of finger movements were obtained using the Mean Absolute Value (MAV) and classified using Support Vector Machines (SVM) and Random Forest (RF) methods. In offline tests, 99.47% accursacy with RF and 98.2% accuracy with SVM were obtained. The RF algorithm with 99.47% accuracy in offline tests was selected and integrated into the embedded system for online tests. In the online tests performed with five volunteers, the system was able to analyze finger movements with an average accuracy of 92.16%, and the commands associated with the finger movements analyzed by the system were sent to the clients with the User Datagram Protocol (UDP), and the related movements were displayed on the output unit interface. The system can work in real-time with a delay of 90 ms and instantaneous movements can be seen visually on the designed output unit interface. This study is an important step in the detection of muscle diseases, the control of EMG-based wearable prosthetic systems, and the design of unmanned vehicles that can be controlled by finger movements.
Subject
Colloid and Surface Chemistry,Physical and Theoretical Chemistry
Reference36 articles.
1. [1] Engin K. O. Ç., Bayat O., Duru D. G. and Duru A. D., “Göz hareketlerine dayalı beyin bilgisayar arayüzü tasarımı”, International Journal of Engineering Research and Development, 12(1): 176-188, (2020).
2. [2] Meyns P., Van de Crommert H. W. A. A., Rijken H., Van Kuppevelt D. H. J. M. and Duysens, J. "Locomotor training with body weight support in SCI: EMG improvement is more optimally expressed at a low testing speed” Spinal cord, 52(12): 887-893, (2014).
3. [3] Fong E. M. and Chung W. Y., “Mobile cloud-computing-based healthcare service by noncontact ECG monitoring” Sensors, 13(12): 16451-16473, (2013).
4. [4] Ocal H., DOĞRU İ. and BARIŞÇI N, “Internet of Things in Smart and Conventional Wearable Healthcare Devices”, JOURNAL OF POLYTECHNIC-POLITEKNIK DERGISI, 22(3), (2020).
5. [5] Aydin E. A., Bay O. F., & Guler I., “P300-based asynchronous brain computer interface for environmental control system”, IEEE journal of biomedical and health informatics, 22(3): 653-663, (2017).