Surface Multi-Purposes Low Power Wireless Electromyography (EMG) system Design

Surface Multi-Purposes Low Power Wireless Electromyography (EMG) system Design Mai S. Mabrouk Biomedical Engineering Misr University for Science and Technology Al Motamyez Distinct 6 October, Egypt Olfat A. Kandil Physical Therapy, Biomechanics Dept. Misr University for Science and Technology Al Motamyez Distinct 6 October, Egypt ABSTRACT The progress in the field of electronics and technology as well as the processing of signals coupled with advance in the use of computer technology has given the opportunity to record and analyze the bio-electric signals from the human body in real time that requires dealing with many challenges according to the nature of the signal and its frequency. This could be up to 1 khz, in addition to the need to transfer data from more than one channel at the same time. Moreover, another challenge is a high sensitivity and low noise measurements of the acquired bio-electric signals which may be tens of micro volts in amplitude. For these reasons, a low power wireless Electromyography (EMG) data transfer system is designed in order to meet these challenging demands. In this work, we are able to develop an EMG analogue signal processing hardware, along with computer based supporting software. In the development of the EMG analogue signal processing hardware, many important issues have been addressed. Some of these issues include noise and artifact problems, as well as the bias DC current. The computer based software enables the user to analyze the collected EMG data and plot them on graphs for visual decision making. The work accomplished in this study enables users to use the surface EMG device for recording EMG signals for various purposes in movement analysis in medical diagnosis, rehabilitation sports medicine and ergonomics. Results revealed that the proposed system transmit and receive the signal without any losing in the information of signals. Keywords Electromyography, wireless, Biceps brachii muscle, surface electrodes, Data acquisition instrument. 1. INTRODUCTION Advances in electronics, signal processing techniques, and in computer technology have made it possible to record and process human bioelectrical signals in real time. Measuring human bioelectrical signals sets some requirements for the measuring system [1]. In most cases, the whole measurement signal bandwidth is needed to be transmitted and because the sampling frequency is normally in the range of 1 khz or even more and several measurement channels may be in use simultaneously, relatively high data transmit rates are needed. Today s low power wireless data transfer techniques are, however, able to meet this challenging demand [2]. Electromyography (EMG) is described as a device for recording the muscle activity that can be used to measure the change in the electrical potential between two points of a muscle using EMG electrodes connected to these two points [3]. It produces electrical signals as a result of muscle contraction that represents the neuromuscular activity. This means that the muscle will generate electrical signal every time it contracts. The amplitude of the electrical signal generated by the muscle activation will vary according to the force applied to move the joint, the velocity of muscle contraction and the angle of muscle pull when it contracts [4][5]. The discovery and development of the EMG signals started early in the 1666 by Francesco Redi, who was the first scientists to start, develop and research the EMG signals. He published a documentary informing that highly specialized muscles of the electrical ray fish generated electricity. After over hundred years later Walsh was able to demonstrate that the muscle tissues of an Ele fish can generate electrical spark. In 1792, A. Galvai published a paper based on electricity being able to initiate muscle contractions. In 1849, Dubios- Raymond found that it is possible to record the electrical activity of a muscle. After over 50 years later Marey was the first to record the electrical signal of muscle activation and he was the one who gave the electrical signal generated by the muscle the term Electromyography [6]. In the 1922, Gasser and Erlanger used an Oscilloscope to show the EMG signal from a muscle. However, due to the nature of the EMG signal and the noise affecting the signal, only rough information of the EMG signal could be seen and analyzed. Between the 1930s and 1950s the detection of the EMG signals improved significantly, since researches and developers started to use improved electrodes for the study of the muscles functions. After 1960 clinician started to use surface EMG electrodes for specific muscle treatments, of these were Hardyck and his research team in 1966. In the 1980 s there was a generation of new Microprocessor based EMG devices that were available at a reduced cost, and easier to use. In the last 15 years, research and development of the EMG signals has resulted in 10

a better understanding of the muscles and EMG signal functionality. The use of EMG during dynamic activities of various velocities necessitates the use of wireless rather than wired EMG [7]. Wireless technologies such as Wi-Fi and Bluetooth have also been incorporated into today s existing EMG equipment to provide the user with extended mobility from the PC on PCbased systems. Acquired EMG signals can now be picked up on the body and sent wirelessly to a PC where it is recorded, processed and analyzed [8, 9]. The electrical EMG signals generated by the muscles similar to other human biopotentials have very low amplitude. The amplifier in EMG device is used to amplify the EMG signals collected from the human muscle to a TTL level (±5 volts) so that it can be read by a computer using the data acquisition instrument (DAQ instrument). It is used to digitalize the analogue EMG signal so it can be read and stored into the computer. Thus, many critical factors must be considered when collecting the EMG signals, for example, noise and artifact problems could distort the signal. Additional DC current could also add offset to the EMG signal. Providing an adequate ground reference is also a significant problem. EMG setup and the need for technical support, experience to deal with sophisticated design and cost, make it optional in the clinical settings. For these reasons, there is a need to develop the setup design to include EMG device as a routine physical diagnostic test. In this work, a system of wireless EMG setup is designed to collect signals from biceps brachii muscle with the required technical support for data processing and analysis. The EMG device is used to collect the EMG signals from the biceps muscle. Moreover, the three surface electrodes are used to collect the EMG signals from the chosen muscle. Despite the relatively long-term use of EMG by investigators in a wide variety of disciplines, little information is available in the literature as to the preferred location of electrode sites. Obviously, for fine wire work the electrode is placed directly in the muscle belly of the muscle of interest. In these cases, similar criteria as used for diagnostic EMG are acceptable. For surface electromyography, however, the decision as to where to put the electrode is much less clear. Some have advocated the use of the site where the muscle can be the most easily stimulated (motor point) as where the maximum amplitude of potentials will be located. Electrode contacts should lie parallel to the muscle fibers not across them. The duration of the electrical events and the velocity of conduction, the electrode center to center separation should be between 2 to 10 mm. Caution is provided to differences in distal versus proximal muscle fibers. Recording contacts should be as large as feasible, meaning that one linear dimension should be at least equal to about half the distance between the pair of electrodes. Bipolar recording contacts should be as similar as possible in size, and impedance. Specific rules apply to the selective recording or rejection of electrical potentials [10]. The chemical electrode transducer is the means by which muscle activity may be detected. A great variety of electrodes are used in EMG but common to all electrodes is a metal electrolyte interface. The electrode is formed of metal, and the electrolyte may be an electrolytic solution or paste, as used with surface electrodes, or the tissue fluids in contact with the embedded electrode. It is at the site of the electrode electrolyte interface that an exchange occurs between the ionic current of the various tissue media and the electron current flow of recording instrumentation. The quality of an electrode as a transducing element depends on the ability of the interface to exchange ions for electrons and vice versa, with equal ease, thus preventing the formation of a charge gradient at the electrode electrolyte interfaces [11]. In the work at hands, the first two electrodes are connected across the biceps muscle to collect the EMG signal of the muscle. The third electrode is located at the elbow region of the same arm of the biceps muscle. This electrode is used to set body reference voltage. The biceps muscle was chosen as a typical superficial and for accessibility of the muscle for project demonstration. Only maximal voluntary isometric task was performed using right dominant biceps. The EMG amplifier must be designed to accommodate two surface EMG electrodes. In order to interpret the information acquired by the EMG amplifier, then DAQ instrument is used to send the data to the computer. The main components of the EMG system are proposed and the data acquisition system is also described in details. 2. METHODS The main idea of this system is to collect the Electromyography (EMG) signals from any muscles in the body such as the Biceps brachii muscle using surface electrodes, using isometric contraction at right angle elbow position and then transmit it to PC via wireless technology. These EMG signals are recorded to be used for movement analysis of the specified muscle. An EMG signal represents the electrical activity of the muscle due to the potential difference following electrode placement when a muscle is contracted. The wireless EMG system is divided into three stages: The first stage is acquiring the EMG signal; the second stage is transmitting the EMG signal using RF wireless. The third stage is receiving it for further analysis. In order to interpret the information acquired by the EMG electrodes, Analogue to digital converter (ADC) is used to digitize acquired EMG signals and Data Acquisition (DAQ) instrument is used to transmit the data to the computer and plot transmitted data for medical diagnosis, figure 1 describes the block diagram of the proposed overall Wireless Electromyography (EMG) system design. Fig 1: Block diagram of the proposed overall Wireless Electromyography (EMG) system design. 11

2.1 Wireless EMG System The EMG signal is acquired by placing conductive elements (electrodes) to the skin surface of the specified muscle. Three silver surface electrodes were used to acquire EMG signals of the biceps muscle. The first electrode is placed on the skin above the middle of the length of the desired muscle. Let's call this the mid muscle electrode. Next, a second electrode is placed before the insertion of the muscle. We'll call this the end muscle electrode. Last, the third electrode is placed on a bony part of the body nearby the muscle group. We'll call this the reference electrode; electrode placement is shown in Figure 2. The basic communications system consists of the following stages: - Transmitter: the sub-system that takes the information signal and processes it prior to transmission. The transmitter modulates the information onto a carrier signal, amplifies the signal and broadcasts it over the channel. The transmitter circuit diagram is shown in figure 3. - Channel: The medium which transports the modulated signal to the receiver. Air acts as the channel for broadcasts like radio. - Receiver: The sub-system that takes in the transmitted signal from the channel and processes it to retrieve the information signal. The receiver must be able to discriminate the signal from other signals which may use the same channel (called tuning), amplify the signal for processing and demodulate (remove the carrier) to retrieve the information. It also then processes the information for reception [12]; the receiver circuit diagram is shown in figure 4. The amplified EMG recordings are fed to the shift amplifier to identify the positive and negative parts of the signal and set them to the standard TTL level (±5 volts) to be transferred to the computer [14]. The amplified EMG recordings are then fed to the phase-locked loop (PLL). The HEF4046B is used as a phase-locked loop circuit that consists of a linear Voltage Controlled Oscillator (VCO) and two different phase comparators with a common signal input amplifiers and a common comparator input [15]. The TX/RX (SHY-J6122TR) 433 MHZ Modules are used for both transmit and receive operation. The two modules come together; one is receiver and the other is transmitter with a transmit power of 1W (10m), operating frequency of 433.92MHZ and operating voltage of 5V. Antenna (TLB 433 2.5N) with a frequency range (433MHz), input Impedance (50Ω), max-power (50W), vertical polarization Vertical, weight (5g) and height (45mm) is used as an electrical device which couples radio waves in free space to an electrical current used by a radio receiver or transmitter. Two first-order low pass filters connected or cascaded together to form a second-order or two-pole filter network is used before the transfer of EMG recordings to the computer. Fig 3: Circuit Diagram of Transmitting stage. Fig 2: Electrode placement. The EMG signal level is too low to be directly captured by a computer which uses the standard TTL level (±5 volts). Therefore, an amplification of the signal is required to increase the signal amplitude up to the TTL level. AD524 amplifier is used because of its high Common Mode Rejection Ratio (CMRR) in addition to; it is used to reject a small level signal noise which is common in both input signals and displaying the output signal as difference in voltage between the two input signals. Also, it is a very high input impedance buffer amplifier required for buffering the input voltage and preventing any voltage drop so that the same voltage at the electrodes appears at the differential inputs without any loss [13]. Fig4: Circuit Diagram of Receiving stage. 12

2.2 Data Acquisition (DAQ) The purpose of data acquisition is to measure an electrical or physical phenomenon such as voltage, current, temperature, pressure, or sound. PC-based data acquisition uses a combination of modular hardware, application software, and a computer to take measurements. While each data acquisition system is defined by its application requirements, every system shares a common goal of acquiring, analyzing, and presenting information. Data acquisition systems incorporate signals, sensors, actuators, signal conditioning, data acquisition devices, and application software. In this work, Data Acquisition (DAQ) is used for acquiring signals from real-world phenomena, digitizing the signals and analyzing, presenting and for saving the EMG data. It offers a significant cost and power reduction compared to a traditional flash analog-to-digital converter (ADC). The DAQ system is described as in figure 5. that can be used for statistical or higher order analysis; figure 6 describes raw EMG signal processing. Fig 6: Raw EMG signal processing. 2.2.2 Normalization The task being studied may also need to be normalized in terms of time. A cyclic activity can be set at 100 N for each cycle. This procedure allows for comparisons of data which may vary slightly in duration. This type of normalization has been used in gait analysis and when carrying loads on the back [16]. While studying shoulder joint load and muscle activity during lifting of a box, arborelius and associates expressed each task as a working cycle ratio, figure 7 describes the normalization processing of EMG signal. Fig 5: The DAQ system. A portable DAQ (NI USB-6008) with a USB connection was used for being simple and low-cost multifunction I/O device from National Instruments that can real time the signal and present it in LabVIEW, buffers the signal, Feeds the circuit by + 5V and used as analog to digital converter. Typically, DAQ boards are installed in a PC with a high speed data bus like the PCI bus. IniLab1008 communicates with a PC via the USB protocol. Depending on the speed of the motherboard of the PC, the maximum data transfers can occur between microprocessor and memory at 20 MHz to 40 MHz's. Sampling frequency and resolution are very important factors in determining the performance of a DAQ card. In this paper, we used a surface electrodes wireless EMG, to get a clear signal and apply several EMG signal processing techniques using LabVIEW such as normalization, integration, Root Mean Square (RMS), frequency spectrum and linear envelop. 2.2.1 Raw Electromyography Signal The raw EMG signal is a random signal obtained from the surface electrodes and then amplified. The raw signal should be monitored for all investigations, because the investigator can pick up major artifacts and eliminate that area or part of the signal. The ergonomist should be aware that on-off information is nominal data but the relative amount of activity is ordinal data. Interpretations that can be made in ergonomic studies are dependent on the amplifier gain and the sensitivity settings of recording instruments. No current standards exist for instrument settings or interpretive rules; therefore, considerable judgment needs to be exercised in evaluating EMG records of raw data. Such a data form is of limited value when findings are to be related to muscle force or fatigue. Thus, the signal is processed to attain a quantitative estimate Fig 7: The normalization processing of EMG signals. 2.2.3 Integration Integration requires full wave rectification followed by filtering and the use of the integrator. Integration uses all parts of the signals and represents the total amount of energy of the signals. The processed signal represents the number of motor units firing, their firing rate, the area of motor unit Integration fails to discriminate between artifacts and motor units. Noise also is a problem when recording low force level contractions. Durations may detect an unacceptably large proportion of noise when investigating low force level contraction [17]. Figure 8 describes the integration processing of EMG signals. 13

there may not be much emphasis forthcoming on this form of analysis; figure 9 describes the frequency spectrum processing of EMG signals. Fig 8: The integration processing of EMG signals. 2.2.4 Root Mean Square (RMS) The root-mean-square (RMS) voltage is the effective value of the quantity of an alternating current. The true RMS value of a myoelectric signal measures the electrical power in the signal. To obtain the RMS value, a ballistic galvanometer, thermocouple, strongly damped voltmeter, or digital computer may be used. A nonlinear detector may be used instead of a linear detector. The RMS signal has immediate relationship to the power spectrum. The curve represents power of the myoelectric signals. The power value is equal to the area under the spectral curve. The signal amplitude is proportional to the square root of the total signal power. The RMS has a linear relationship with tension for a brief isometric contraction; figure 8 describes the RMS processing of EMG signal. Fig 8: The RMS processing of EMG signals. 2.2.5 Frequency Spectrum The square of the magnitude of fast Fourier transform (FFT) is used for analysis [18]. The center frequency is not affected by force of contraction except at low level where it increases with increase in force because the use of spectral changes in the myoelectric signal as a valid indicator of muscle fatigue has been questioned. Frequency analysis based on zero crossings is used to study shoulder and neck disorders in assembly line workers. The results are promising and suggest that the technique is valuable for ergonomic studies at the workplace give the availability of other techniques however Fig 9: The frequency spectrum processing of EMG signals. 2.2.6 Linear Envelope A linear envelope can be used to provide an envelope that represents a profile of the myoelectric activity of the muscle over time. 2.3 Implementation A wireless EMG system is proposed to acquire and analyze EMG signals in real-time. A good visualization and analysis of these signal is essential using a successful software package. This work is written and implemented in the LabVIEW language and has been tested on the WINDOWS platform with LabVIEW. Our GUI provides a number of functions that can be used to create graphic output. A simple graphical user interface (GUI) has been developed by using LabVIEW. It aids the usage of a visual programming language that suits very well for hardware integration, and suits for general purpose programming; a snap shot of the GUI is shown in figure 6. The graphical user interface (GUI) allows the users to identify patient ID and select bandpass filters as required. 3. RESULTS AND DISCUSSION The Electromyogram (EMG), based on changes in amplitude and frequency, can be qualified and used to classify the electrical activity level that produces a certain muscular tension. The change in the myoelectric signal is based on the requirement and firing rate of motor units within the muscle, in general, as more force is needed, more motor units are required, and the motor units already firing increase their frequency of firing. This general reaction, however, is not exactly the same for every muscle. The interpretation of the changes in requirement and changes in firing rate can provide information concerning the muscle's level of force or its level of fatigue [19]. The information in this work presents a variety of ways by which the ergonomist may analyze or subsequently interpret myoelectric activity for different workplace needs. Many different methods are used to reduce the data contained in the electrical signal and to present it in numerical form. 14

Which method to use depends on the reason for which the information is needed, that is, the purpose of the study. The interpretation of the EMG signal plays an important role in determining the relationship of muscle activity to task performance. The most basic information obtained from a myoelectric signal is: 1) whether or not the muscle is active on/off and 2) the relative amount of activity of the muscle. By using the appropriate process of normalization, a reasonable estimate of muscle function can be obtained by the ergonomist [20]. This information can be combined with an observation system or simply an event marker of some type to determine: 1) The time at which the muscle is active. 2) The time at which a peak of activity occurs. 3) The pattern of muscle activity during a movement between agonist and antagonist. The raw, or unprocessed, EMG signal is the basis of all methods of interpreting the myoelectric activity from muscles. The ergonomist should monitor the raw signal, even though other signal processing may be used, so that artifacts can be detected and controlled as necessary for spike counting or whatever. In the past, probably the most common way to interpret EMG was by visual inspection of the raw signal with training, experience, and the use of multiple gains and oscilloscope sweep vibration, the observer should be able to evaluate the raw EMG signal visually and effectively. The observer should be able to identify when the raw signal indicates that a muscle is active and when it is relaxed. The relative amount of activity may be classified either qualitative by words, such as nil, negligible, slight, moderate, marked, or very marked, or quantitative by numerical values, such as 0-5, with 0 being no activity and 5 being maximal activity. Figure 7 shows the GUI of the proposed Wireless EMG system with recorded and processed EMG data. Fig 10: The GUI of the proposed Wireless EMG system with recorded and processed EMG data. 4. CONCLUSIONS The monitoring of vital physiological signals has proven to be one of the most efficient ways for continuous and remote tracking of the health status of patients or performance. The aim of this work is to design a wireless EMG device to help eliminate the restrictions caused by lead wires in conventional systems by permitting the acquisition and wireless transmission of a signal from sensor to recorder. This work presents the implementation of the surface EMG device. The device is implemented in three stages; analogue signal processing, DAQ and computer EMG (GUI) Capture software. The EMG analogue signal processing stage is designed to amplify EMG signals from one biceps muscle. The EMG analogue signal processing stage consists of five major components: surface EMG electrodes, Instrumentation amplifier, Shift amplifier, low pass amplifier and full transmitter and receiver wireless. The surface EMG electrodes are used as input sensors to the EMG amplifier. The amplifier is a differential instrumental amplifier which amplifies the weak and noisy EMG signals to a stable TTL level for further processing. The low pass filters are used to select only the bandwidth of signals that is of interest for the application. The wireless used to transmit and receive the signal without any losing the valuable information of signals. The DAQ stage is used to digitalize the EMG signals using the built in 12bit A/D converter. The second function of the DAQ is to transmit the digitalized EMG signal using the USB Serial communication module built in the DAQ. The EMG capture GUI program LabVIEW is developed to read the save EMG data.. The program provides necessary functions such as reading the saved EMG data filter the EMG data and analyze it. The program also provides plotting function to be used to plot the processed EMG data on a graph. This work represent phase I of the muscle recognition scenario during dynamic activity. Phase II of this work will deal with agonist /antagonist relationship. 5. REFERENCES [1] A. Cohen, "Biomedical Signals: Origin and Dynamic Characteristics; Frequency Domain Analysis," in the Biomedical Engineering Handbook, J. Bronzino Ed. Boca Raton, CRC Press, 1995. [2] Antti T. Vehkaoja, Jarmo A. Verho, Merja M. Puurtinen, Niina M. Nöjd, Jukka O. Lekkala, and Jari A. Hyttinen, Member, IEEE, Wireless Head Cap for EOG and Facial EMG Measurements, Proceedings of the 2005 IEEE Engineering in Medicine and Biology 27th Annual Conference Shanghai, China, September 1-4, 2005 [3] Webster, John G. (1998). Medical Instrumentation: Application and Design. [4] Crosbie, J. H., Lennon, S., Basford, J. R., & McDonough, S. M, Virtual reality in stroke rehabilitation: Still more virtual than real. Disability and Rehabilitation, 29(14), 1139 1146, 2007. [5] Y.T. Yu, M.F. Lau, "A comparison of MC/DC, MUMCUT and several other coverage criteria for logical decisions", Journal of Systems and Software, 2005. [6] Abbink, V. D. B., Glas, V. D., Detection of onset and termination of muscle activity in surface electromyograms, Journal of Oral Rehabilitation, 25(5), 365-369. 1998. [7] Hassoun, M. H., Wang, C., Spitzer, A. R., NNERVE: Neural network extraction of repetitive vectors for electromyography II, performance analysis. IEEE Transactions on Biomedical Engineering, 41(11), 1053-1061, 1994. [8] Vehkaoja, A., Lekkala, J., Wearable Wireless Biopotential Measurement Device, in Proceedings of 15

the 26th Annual Engineering in Medicine and Biology Conference, San Francisco, pp. 2177 2179, 2004. [9] [Hughes, E., Bell, A., A Wireless Surface Electromyography System, IEEE Southeast Conference Proceedings, 253, 2007. [10] Eisenhower, W., McMichael, S., Analog control with surface electromyography, Electrical and Computer Engineering. University of Maryland, College Park, 2005. [11] B. D. Farnsworth, R. J. Triolo and D. J. Young, Wireless Implantable EMG Sensing Microsystem, IEEE Sensors Conference, pp. 1245-1248, 2008. [12] Noraxon, I. (2010). Zero-Wire - True Wireless EMG Retrieved 4 April 2010. [13] M. Suster, J. Guo, N. Chaimanonart, W. H. Ko, and D. J. Young, A Wireless Strain Sensing Microsystem with External RF Powering and Two-Channel Data Telemetry Capability, ISSCC, pp. 380-381, 2007. [14] Op Amps for Everyone, Ron Mancini (Ed.), Chapter 12, Texas Instruments Literature Number SLOD006. [15] N. Chaimanonart, M. Suster, and D. J. Young, Two- Channel Data Telemetry with Remote RF Powering for High-Performance Wireless MEMS Strain Sensing Applications, IEEE Sensors Conference, pp. 285-288, 2005. [16] Pelland, L., McKinley, P., A pattern recognition technique to characterize the differential modulation of co-activating muscles at the performer/environment interface, Journal of Electromyography and Kinesiology, 14, 539-554, 2004. [17] Raez, M. B. I., Hussain, S., Mohd-Yasin, F, Techniques of EMG signal analysis: Detection, processing, classification and applications, Biological Procedures Online, 8, 11-35, 2006. [18] De Michele, G., Sello, S., Carboncini, M. C., Rossi, B., & Strambi, S., Cross-correlation time-frequency analysis for multiple EMG signals in Parkinson's diease: A wavelet approach, Medical Engineering and Physics, 25(5), 361-369, 2003. [19] Jaskolska A, Kisiel-Sajewicz K, Brzenczek-Owczarzak W, Yue G H. c, Jaskolski A, EMG and MMG of agonist and antagonist muscles as a function of age and joint angle, Journal of Electromyography and Kinesiology 16, 89 102, 2006. [20] Evetovich T K, Naumann J, Conley D S, and Todd J B, Effect of Static Stretching of the Biceps Brachii on Torque, Electromyography, and Mechanomyography During Concentric Isokinetic Muscle Actions, Journal of Strength and Conditioning Research, 17(3), 484 488, 2003. 16