SEMG ACQUISITION CIRCUIT AND DISPLAY ON COMPUTER

Nguyen Anh Duy Le Thi Yen Bui Dinh Hieu HANOI UNIVERSITY OF SCIENCE AND TECHNOLOGY

Contents
I. II. III. A. Introduction .................................................................................................................... 3 Specifications ................................................................................................................. 3 Background Information ............................................................................................. 3 Electromyography ...................................................................................................... 3

Electromyography (EMG) is an experimental technique concerned with the development, recording and analysis of myoelectric signals. Myoelectric signals are formed by physiological variations in the state of muscle fiber membrane. ......................................... 3 B. C. IV. A. 1. 2. 3. 4. B. 1. 2. V. Noise and artifact ....................................................................................................... 3 Disadvantages ........................................................................................................ 4 Choosing elements ..................................................................................................... 4 Acquisition Circuit ....................................................................................................... 4 Preamplifier............................................................................................................. 4 Bandpass Filter ....................................................................................................... 6 Amplifier with Gain Adjustment ............................................................................. 13 A/D converter ........................................................................................................ 13 EMG Signal Processing on LabVIEW ....................................................................... 14 Environment.......................................................................................................... 14 EMG Signal Processing ........................................................................................ 15

Evaluation and Future Work ......................................................................................... 17

Caption Menu
Figure 1: INA128 Schematic ................................................................................................. 5 Figure 2: INA128 Simplification Form .................................................................................... 5 Figure 3: Choosing Filter Type in Filter Pro ........................................................................... 7 Figure 4: Illustration for setting up filter specification ............................................................. 8 Figure 5: Choosing filter respond type ................................................................................... 9 Figure 6: Choosing filter topology type ................................................................................ 10 Figure 7: Sallen - Key Highpass filter .................................................................................. 10 Figure 8: Sallen - Key Lowpass filter ................................................................................... 10 Figure 9: EMG highpass filter .............................................................................................. 11 Figure 10: EMG lowpass filter ............................................................................................. 11 Figure 11: TL072 Schematic ............................................................................................... 12 Figure 12: NI USB 6008 ...................................................................................................... 14 Figure 13: VI block for LabVIEW processing ....................................................................... 16 Figure 14: Reading and Logging File Block ......................................................................... 16 Figure 15: Visual Display Acquisition .................................................................................. 17 Figure 16: Reading File Visual Display ................................................................................ 17 Figure 17: Some recorded samples .................................................................................... 18

Table Menu
Table 1: Some parameters of INA128 ................................................................................... 5 Table 2: Some parameter of TL072CP ................................................................................ 13

I.

Introduction
Many people are suffering from neuromuscular diseases such as stroke, Parkinson, brain or spinal cord injury leading to complete or partial palsy. Those people need rehabilitation process in order to recover normal functions of their body. Nowadays, technology innovation leads a major growth in the high-tech rehabilitation applications for disabilities. Many new technologies were applied for creating smart, new approaches in rehabilitation. The robots would be able to carry out exercise that a therapist would do for the patients and some are even not so easy to be carried out by a human being. The exoskeletons are wearable robots exhibiting a close cognitive and physical interaction with the human user. These are rigid robotic exoskeleton structures that typically operate alongside human limbs. A robotic exoskeleton system (or an exoskeleton robot) is a novel man-machine intelligent system which fully merges the human intelligence and machine power. The robotic exoskeleton transmits torques to the human joints from the actuators through its links. The technology in this field has acquired a rapid development in recent years. An exoskeleton can be controlled by different types of parameters. The suitable approach is using electromyography (EMG) signal because it provides rich motor control information from which the users intention of motion can be detected. Therefore, in this thesis we focus on the acquiring sEMG and transmit it to computer for processing and displaying, which is just a small part of the problem.

II. III.

Specifications Background Information

A. Electromyography
Electromyography (EMG) is an experimental technique concerned with the development, recording and analysis of myoelectric signals. Myoelectric signals are formed by physiological variations in the state of muscle fiber membrane.

B.

Noise and artifact

There are three sources of noise affecting on EMG acquisition which are ambient noise, transducer noise, and motion artifact. Ambient noise is generated by electromagnetic devices surrounding the measuring place such as computer, power lines, television, motors, fluorescent lamps, light bulb, etc. They have the frequency ranging between 0 to 1000Hz. However the dominant frequency component is 50 or 60 Hz which are the power line frequency

Transducer noise is the noise generated at the electrode to skin junction. It can be categorized into DC voltage potential and AC voltage potential. The DC voltage potential results from the different impedance levels between the skin and the electrode as well as chemical reactions that occur between the electrode and the conducting gel. The AC voltage potential results from the changing of this impedance level. This type of noise should be reduced by using appropriate materials and consistently applying procedures. The last type of noise is motion artifacts which could be from the skin-electrode surface of from the movement of the cable connecting the electrode to analog circuit. Those noises have most of their energy in the frequency range from 9 20 Hz. Hence, they could be eliminated by using the highpass filter with cutoff frequency above 20 Hz. In EMG measurements we might, from time to time, also experience ECG/EKG artifacts (signal from the heart muscle). These artifacts are present due to the fact that a heartbeat causes a relatively high electric field in the human body. Depending on the location of the measurement, this is present in an EMG measurement or not.

C.

Disadvantages
Obtaining the same EMG signal for the same motion is difficult even with the same person since the signal is biologically generated The activity level of each muscle and the way they are used for certain motion is different among individuals Various muscles are involved in a joint motion The role of each muscle for a certain motion varies with joint angles The activity levels of some muscle are affected by the movement of the other joints

IV. Choosing elements

A. Acquisition Circuit
1. Preamplifier The general idea of the differential amplifier stage is in figure. Basically, it uses an o-amp to amplify the voltage difference between two inputs from a pair of surface electrodes attached on the muscle bundle. On each input, there are not only the EMG signals, but also noises. Since the similar noises on the two inputs will be reduced considerably by the differential amplifier, only the voltage different passes. The amount of noise deduced from the desired signals is determined by the common mode rejection ratio (CMRR) of the opamp. The higher CMRR, the better noise rejection. Hence the op-amp chosen should have high CMRR, high input impedance, low output impedance and good adjustable gain. For those

requirements, INA128 which is widely used in biomedical applications is a good choice.

Figure 1: INA128 Schematic

Figure 2: INA128 Simplification Form

INA128 is low power, general purpose instrumentation amplifiers offering excellent accuracy. The versatile 3 op-amp design and small size make them ideal for a wide range of applications. Currentfeedback input circuitry provides wide bandwidth even at high gain (200 Khz at G = 100). A single external resistor RG sets any gain from 1 to 10000 by the equation:

Table 1: Some parameters of INA128

Parameter Gain CMRR Power Supply Voltage Power Supply Current Operating

Min 1

Typ

Max 10000

120 dB at G 100 2.25V 15V 700 -40oC 18V 750 125oC

5

Temperature Range Those parameter prove that INA 128 fulfils our requirement. Following is the schematic of the differential amplifier stage in our circuit The gain is determined by the resistor connected to pins 1 and 8. For RG = 120 , G =418. The EMG signals of amplitude 0.1 - 5mV will be amplified to 40mV 2 V which is possible for acquisition or will be adjusted by an amplifying stage following. 2. Bandpass Filter The next stage is a bandpass filter for removing the undesired signals outside the range of 20 500 Hz since most of the noteworthy EMG signals lie inside that range. For better response, the bandpass filter is assembled form a low pass filter and a high pass filter in series. Each of those is designed by Filter Pro Desktop a free active filter design application by National Instrument. Here are the steps in designing filters using this software: Step 1: At start, it asks to choose the type of filter, lets say Lowpass (figure 3)

Figure 3: Choosing Filter Type in Filter Pro

Figure 4: Illustration for setting up filter specification

Step 2: Set the filters specifications (passband fre quency 500 Hz, set fixed filter order 6, others left as default) (figure). The 4 order filter was also implemented, however, its frequency response is not good enough, so we increased the order to 6, It is sufficient and should not be increased more making the hardware too much cumbersome. Step 3: Choose the filter response type As we know, there is the trade-off between the sharpness of the cut-off region and the ripple in the passband. We do not need the sharpness too high, as in Chebyshev case where there is much ripple, because we need to well keep signals in the pass band. Besides, the signals will be filter further digitally on Labview. That is why Butterworth filter, as showed in the figure which has the flat gain in passband with quite good cut-off sharpness, is the good choice for our requirement. Step 4: The last step is choosing the filter topology The Sallen-Key filter is chosen since it is a very popular active filter which is particularly valued for its simplicity and can be cascaded to form higher order filter. This filter type is a two-pole filter topology, introduced by R. P. Sallen and E. L Key of MIT Lincoln Laboratory in 1955. It is degenerated frin Voltage Controlled Voltage Source (VCVS),
8

and the available in low pass, high pass, bandpass, and notch versions.

Figure 5: Choosing filter respond type

Figure 6: Choosing filter topology type

Figure 7: Sallen - Key Highpass filter

Figure 8: Sallen - Key Lowpass filter

10

Its cut-off frequency can be defined:

We can calculate it manually; however Filter Pro Desktop help us to cascade stages, change component easily, and observe the result immediately. The highpass filter is done in the same process. The following figures show the complete filter schematic and its frequency response. The op-amp chosen for this filter stage is TL072Cp which is low noise version with low input bias and offset currents and fast slew rate. The low harmonic distortion and low noise make the TL072x series ideally suited for high fidelity applications like ours.

Figure 9: EMG highpass filter

11

Figure 11: TL072 Schematic

12

Table 2: Some parameter of TL072CP

Parameters CMRR (Min) (dB) Operating Temperature Range oC Pin/package Power Supply

Value 70 -40 to 85 8 DIP, 8 SOIC 18 V maximum

3.

Amplifier with Gain Adjustment This amplifier stage is simply the invert amplifier with the adjustable gain from 1 to 5. The gain is given by

Adjusting the gain in case the input EMG signals is too small will put the signals into a proper range for the A/D converter. In this stage TL072CP is also used 4. A/D converter To record the EMG signals from the amplifier circuit to computer for further processing, an A/D converter, which converts the analog signals to digital ones, is needed. Using the A/D converter, we should take care of its clock speed, the resolution bit rate, the number of analog input/output channels, or the software that comes with it Since, LabVIEW is chosen for the computer-based signal processing in next part, the Data Acquisition Card (DAQ) also from National Instrument which is fully sipported for LabVIEW, should be an excellent choice. The DAQ being used is the student version, NI USB6008. It is multifunctional low-cost DAQ that can be connected directly to a computer via USB port.

13

Figure 12: NI USB 6008

This is some parameter of NI USB 6008: 8 analog inputs (12bit, 10kS/s) 2 analog outputs (12bit, 150S/s); 12digital I/o; 32 bit counter Bus powered for high mobility; built in signal connectivity OEM version available Compatible with LabVIEW, LabWindows/CVI and Measurement Studio for Visual Studio .NET NI DAQ max driver software and NI LabVIEW Signal Express LE interactive data-logging software

This card is sufficient for our system because it has the maximum sampling frequency 10 KHz, 4 pairs of analog inputs for acquiring at most four sEMG channels.

B.

EMG Signal Processing on LabVIEW

1. Environment LabVIEW (Laboratory Virtual Instrument Engieering Workbench), product of National Instrumentation, is a higle productive development environment that engineers and scientists use for graphical programming and unprecedented hardware integration to rapidly design and deploy measurement and control systems. Within this flexible platform, engineers scale from design to test and from small to large systems while reusing IP and refining their processes to achieve maximum performance.

14

LabVIEW software is ideal for any measurement or control system, and the heart of the NI design platform. Integrating all the tools that engineers and scientists need to build a wide range of applications in dramatically less time, LabVIEW is a development environment for problem solving, accelerated productivity, and continual innovation. LabVIEW is an excellent environment to make the graphical user interface. It contains a comprehensive collection of drag-and-drop controls and indicators to help users quickly and easily create user interfaces for application and effectively visualize results without integrating third party components or building views from scratch. Power user can also customize the built-in controls via the Control Editor and programmatically control user interface elements to create highly customized user experiences. 2. EMG Signal Processing The EMG signals acquired through the A/D converter will be process on LabVIEW which does digital filtering, feature extraction, spectral analysis, and data logging and playing. The digital filter contains a 50 Hz notch filter and a 20 500 Hz band pass filter. The feature extraction is for getting the meaningful characteristics of EMG signals which relating to the arms movements. The spectral measurement is just for observing, but has not been used as a feature for controlling. The logger and display is for writing the measured signals to a file or opening that file for displaying.
Start

Acquired by DAQ

Notch Filter (50 Hz)

Bandpass Filter (20 500 Hz) Spectrum Measurement RMS measurement

15

Display

Display

Figure 13: VI block for LabVIEW processing

Figure 14: Reading and Logging File Block

16

Figure 15: Visual Display Acquisition

Figure 16: Reading File Visual Display

V.

Evaluation and Future Work

More than 100 sEMG samples were recorded, however 60% has very large noise and cant be used. The circuit is not working stably. The reasons may cause by some of the connections between the acquisition circuit and NI USB, NI USB and Laptop or power supply since there are 4 phone plug to connect with the circuit and all 4 plugs dont working well. If we have more time we will develop more about the processing by LabVIEW to get cleared signal. Then build full system with an exoskeleton and try to control the skeleton via EMG signal acquired Below are some figure of the recorded samples
17

Figure 17: Some recorded samples

18