Signal Acquisition and Time–Frequency Perspective of EMG Signal-based Systems and Applications

Abstract

AbstractThe last few decades have emerged as a remarkable era for exploring and employing electromyography (EMG) signals and their attributes in various applications such as clinical assessment and rehabilitation engineering. An EMG signal-based system encapsulates different domains of signal acquisition and processing, statistical analysis, and control systems in a single framework. This survey attempts to highlight and distinguish the time- and frequency-based signal processing according to the applications of EMG signals. When EMG signals are used for clinical assessment, time–frequency analysis involves transforming the signals in different domains and extracting useful physiological information. On the other hand, the concept of time and frequency deals with extracting time, frequency, or time–frequency-based features when EMG signals are used for pattern recognition-based control applications such as robotics and augmented reality. It is often very difficult and confusing to distinguish and establish a clear understanding between these domains reported in various literature. Hence, this study first presents different signal acquisition systems and pre-processing techniques, followed by comprehending the concepts in time, frequency, and time–frequency-based approaches based on the applications. Next, the review of various post-processing techniques, different feature extraction routines, and a survey of different classifiers used in the pattern recognition step is done. The work concludes with a study of innovative applications of EMG signals reported in recent years, provides an overview of EMG signal-based limb prosthetics, and suggests a few futuristic research ideas.KEYWORDS: Biomedical engineeringData acquisitionElectromyographyFeature extractionPattern classificationProstheticsSignal processing Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationNotes on contributorsAnil SharmaAnil Sharma received a BTech degree in electronics and communication and an MTech degree in mechatronics engineering. Currently, he is pursuing PhD in the electronics and communication engineering department from Malaviya National Institute of Technology (MNIT), Jaipur. His current research interest includes biomedical signal processing, EMG signal acquisition and control, machine learning, and robotics. Corresponding author. Email: 2020rec9510@mnit.ac.inIla SharmaIla Sharma received a PhD degree in electronic and communication engineering from the PDPM Indian Institute of Information Technology Design and Manufacturing Jabalpur, Jabalpur, India. Currently, she is an assistant professor with the electronics and communication engineering department at Malaviya National Institute of Technology (MNIT), Jaipur, India. Her current research interests includemulti-ate filter banks, digital signal processing, multiplier-less filters and filter banks, wireless communication, and cognitive radio. Email: ila.ece@mnit.ac.inAnil KumarAnil Kumar received a BE degree in electronic and telecommunication engineering from the Army Institute of Technology, Pune University, Pune, India, in 2002 and the MTech and PhD degrees in electronic and telecommunication engineering from IIT Roorkee, Roorkee, India, in 2006 and 2010, respectively. He is an assistant professor at the electronic and communication engineering department, Indian Institute of Information Technology, Design and Manufacturing, Jabalpur, India. He is currently a visiting researcher with the Gwangju Institute of Science and Technology, Gwangju, South Korea. His current research interests include the design of digital filters and filter banks, biomedical signal processing, image processing, and speech processing. Email: anilk@iiitdmj.ac.in