projekt „issnb“ niš, september 2007- 1 - daad deutscher akademischer austausch dienst german...

Projekt „ISSNB“

Niš, September 2007 - 1 -

DAADDeutscher Akademischer Austausch DienstGerman Academic Exchange Service

Design of ADC Module and a MATLAB Toolbox for Mechanical Myo-tonometry Signals Processing,

Analyzing and Visualization

Galidia Ivanova Petrova, Grisha Valentinov Spasov, Ivan Vladimirov Rachev

TECHNICAL UNIVERSITY OF SOFIA, PLOVDIV BRANCH

Department of Electronics and Department of Computer systems and technologies

Projekt „ISSNB“



Mechanical myo-tonometry (MMYOTON) is a method which performs biomechanical diagnostics of functional state of skeletal muscles. The principle of the MMYOTON lies in using of an acceleration probe to record the reaction of the peripheral skeletal muscle or its part to the mechanical impact and the following analysis of the resulting signal.

The investigated tissue responds to the mechanical impact with damped oscillations.

From this signal several signal parameters could be determined and the characteristic myo-tonometry parameters (muscles frequency, elasticity and stiffness) could be calculated.

The review of the references shows that MMYOTON is a promising method for development of new research of skeletal muscles and their functional state.

Projekt „ISSNB“



Fig.1. Waveforms of Displacement (S), Velocity (V) and Acceleration (A) signals presenting muscle reaction to a short mechanical impact.

The acceleration value of the first period of oscillations, calculated from the oscillation graph, characterizes the deformation of the tissue. The data of the next oscillation period provided the basis for calculating the oscillation

frequency of the tissue (f). (1) ]Hz[

T

1f

Projekt „ISSNB“



The logarithmic decrement (Θ) of damping was calculated according to following relation:(2) The logarithmic decrement of the damped

oscillations characterizes tissue elasticity.

Stiffness (C) reflects the resistance of tissue to the force that changes its shape and it was calculated by:

(3)

The tissue tone is a biomechanical property of the tissue which can be characterized by two main parameters: stiffness and elasticity.

This work describe the main requirements for design and realization a special purpose ADC module for processing and discretization of mechanical myo-tonometry signals and design and operation of a MATLAB toolbox for further processing, analyzing and visualization of these signals.

5a

3aln

]m/N[l

a.mC max

Projekt „ISSNB“



The presented ADC module performs: amplification; filtering; discretization of the three analogue signals, registered by the front-end module MYOTONUS; for a defined period of time; to transfer data to a PC; save data in a file for further visualization and digital processing.

Characteristics of the registered signals are: The frequency of oscillations is in the range of several Hz to 30Hz; The amplitudes of the signals in the input of the module are in the range from -10mV to + 10mV. The period for registration of myo-tonometry signals is approximately 11 sec. The functional block diagram of ADC module for mechanical myo-tonometry signals processing is illustrated on Fig. 2.

Projekt „ISSNB“



Inpu

t s

50 HzStop-band filter

InputAmplifier

Amplification with dc-level

shifting

Inpu

t v


InputAmplifier


shifting

Inpu

t a


InputAmplifier


shifting

Power Supply for +/- 5V

Mic

ro-

cont

rolle

r

MAX 232

RS 232

Fig.2. Functional block diagram of presented ADC module for mechanical myo-tonometry signals processing.

Manual “Start” Button

Projekt „ISSNB“



Every analogue channel consists of three stages: Input amplifier; 50Hz stop-band filter; additional amplification stage with dc-level shifting. The whole gain of amplification in every channel is 200, distributed between input amplifier and the third stage. To position the signals in the middle of input range of AD-converter a dc-level is added to the amplified signal in the third stage of the three analogue channels. This dc-level does not represent the real physical situation for recorded signals (muscle displacement, velocity and acceleration) and has no relation to the interested characteristic muscle parameters.

The choice of used microcontroller (ATMega8) is based on several requirements: low power consumption; internal AD-converter (at least 3-channels); high efficiency; embedded hardware for serial communication; available tools (KIT) for design and programming.

Projekt „ISSNB“



The algorithm of work for microcontroller is divided into two parts. The main part is responsible for:

initial initialization of microcontroller after switching power “ON”; watching for the signal from manual “Start” button; discretization of the signals, coming from the three analogue channels to the AD-converter inputs, with frequency of 675Hz; data conversion into ASCII code; asynchronous serial data transfer, by means of embedded USART controller - (the speed of data transfer is 115 200kbs and the format of data is (8-1-N)). The AD-conversion is with 10-bit accuracy, thus every measured value is represented and transferred to the PC by five bytes – four data bytes and one for symbol “CR”.

The other part of algorithm is dealing with timers. It is responsible for: setting the frequency of discretization to 675Hz; software setting of a flag for start of new measurement; counting the time for the whole period of registration of myo-tonometry signals to be approx. 11sec.

Projekt „ISSNB“



In order to perform digital processing and analyzing of recorded myo-thonometry signals, a special purpose MATLAB Toolbox was designed.

The proposed MATLAB Toolbox is designed to work together with special purpose external ADC module for MMYOTON signals processing and to give convenience to medical doctors in calculation the characteristic muscle parameters, archiving the results and further statistical processing.

It is designed to overcome the drawbacks in a simple arrangement, where for registration of muscle response the conventional three-channel ECG instrument is used and determination and processing of signals parameters are performed manually, which is very time consuming, boring and not enough accurate.

The main requirements for the design of the toolbox could be specified as follow:• to be able to read the multiplexed digital data from ADC module via RS-232 interface;• to be able to save the received data in file and read them later;• to visualized data, received (raw) or saved in a file;

Projekt „ISSNB“



• to be able to apply digital filtration on received signals, when it is necessary in order to reduce the influence of internal and external interferences;• to be able to determine the specific parameters of the recorded signals in purpose to calculate the characteristic muscle parameters; • to have a user friendly interface, because it will be used by medical doctors.

The Toolbox window, shown on Fig. 3, contains three panels. The first one (marked by 1 on the figure) is designed to present text messages and calculated results. The second panel (marked by 2 on the figure) is designed for graphical visualization of the signals. The third panel (marked by 3 on the figure) contains tools (buttons and sliders) for optional signal processing.

After process of acquisition signals are visualized on the graphical panel and the program suggests two possible options: to apply filtration, if the doctor considers it is necessary, or to go for calculation of the results.

Options for digital filtration:• a stop-band filter for 50Hz;• a low-pass filter with cut-off frequency (Fpb), which could be changed in a range of 50 – 120 Hz by the corresponding slider tool.

Projekt „ISSNB“



Fig.3. Graphical interface of designed MATLAB Toolbox

Projekt „ISSNB“



Before going to the procedure for calculation of the characteristic muscle parameters, it is obligatory to select a preferred area from the whole signal, called ‘region of interest’ (ROI), wherefrom the signals parameters will be determined.

The shape of the recorded signals is not determined and depends on the experiment conditions and the functional state of the investigated muscle.

It is possible the results of the procedure for signals parameters determination not to be satisfactory and acceptable for the medical doctor. This was the reason for inserting the markers on the signals, as it is illustrated in fig. 4. Then, it is the doctor decision how to proceed further: to select another ROI or to apply filtration.

The results for the calculated values of the characteristic muscle parameters

are displayed in the text panel.

Projekt „ISSNB“



Fig.4. An example for signal parameters determination and calculation of f, C and Theta.

Projekt „ISSNB“



CONCLUSIONS

The designed and realized ADC module and MATLAB Toolbox are given for approbation from medical doctors in Department of Physiology in Medical University of Plovdiv. After collecting enough data and having impressions about the module and program possibilities they could be developed with some new features in accordance with the doctors’ recommendations.

Another area of application of ADC module and the Toolbox is in the educational process for the students in Medical University of Plovdiv and Technical University. They are an useful tool for demonstration of mechanical reaction of skeletal muscles for students in Medicine, from one side, and a good example about the principles, design and application of computer-based systems in processing of biomedical signals for the students in Engineering field, from the other.

The designed Toolbox is easily integrated with other Microsoft programs, which is useful for data archiving and for presentations of the research results.

projekt „issnb“ niš, september 2007- 1 - daad deutscher akademischer austausch dienst german...

Documents

tissue elasticity

projekt issnb ni

registered signals

analogue signals

resistance of tissue

tissue tone

investigated tissue

frequency of oscillations