Graphical Programming based Biomedical Signal Acquisition and
Contents
1. average M A filter is commonly used The general form of an MA filter is 1 s n D bali where x and y are the input and output of the filter respectively The b values are the filter coefficients or tap weights k 0 1 2 N where N is the order of the filter The effect of division by the number of samples used 1 is included in the values of the filter coefficients Applying the z transform we get the transfer function A z of the filter as z X where X z and Y z are the z transforms of x n and y n respectively A simple MA filter for filtering noise is the von Hann or Hanning filter 2 given by N bz b b t tbyz 2 k 0 H z y n 3 The Analog Input Multipoint Block that was used in the lv n 2x n 1 x n 2 Issue 4 Volume 1 2007 319 above block diagram communicates with the data acquisition system and makes the acquisition of an external signal We use an analog channel with following characteristics the name of the channel Ecgl the channel is programmed for signal input Reference Single Ended unipolar signal the input range iS 5mV 5mV the number of samples inside one acquisition is 100 the sampling frequency is 200Hz Fig 5 Part of electrocardiogram acquisition and processing The first three characteristics are appointed by MAX that is capable to communicate to any National Instruments data acquisition card The last two properties ar2. 5 Time s Fig 18 Abdominal ECG and fetal ECG extraction of the ECG signal The noise may be complex stochastic processes within a wideband so you cannot remove them by using traditional digital filters To remove the wideband noises you can use the Wavelet Denoise Express VI asin Figure 19 a b Wavelet Denoise Fig 19 a ECG Wavelet Denoise Express V irtual Instrument Express VI first decomposes the ECG signal into several Abdominal ECG 0 6 0 4 l l 1 3 ais 4 o Amplitude y 2 0 0 2 l l l l l i 4 5 5 5 5 6 6 5 7 7 5 Time s Fetal ECG O1 l l l l K i 4 4 5 5 5 5 6 6 5 7 7 5 5D a Amplitude oO ao on 3 Time s Fig 19 b ECG Wavelet Denoise Express V irtual Instrument and characteristic window for W avelet D enoise subbands by applying the wavelet transform 3 6 and then modifies each wavelet coefficient by applying a threshold or shrinkage function and finally reconstructs the denoised signal 324 INTERNATIONAL JOURNAL OF CIRCUITS SYSTEMS AND SIGNAL PROCESSING The data source can be from terminal or from file From terminal specifies that the virtual instrument reads data from the block diagram input From file specifies that the virtual instrument reads data from a file The virtual instrument can read data from waveform WAV or TXT files The virtual instrument processes the data from WAV files as a waveform data type The valid format of a
3. By performing JCA you can obtain the estimation of S which means you can obtain both maternal ECG and fetal ECG The ASPT provides the TSA Time Sample Analysis ndependent Component Analysis VI with which you can easily build an FHR Fetal Heart Rate extraction application as shown in Fig 17 In Fig 17 8 channel ECG signals acquired by the 8 channel system mentioned before are the inputs of the ICA function and the fetal ECG signal can be obtained from the output as one of the independent components Fig 18 shows 1 channel maternal abdominal ECG 8 channel ECGs Abdominal ECG pe ee EGE EE E a Ei F azeling emong Denoist i No T5A Independent Component Analysis wi i Remove Denoise omponent Index Ez extract Fetal ECG as an independent component Fig 17 ICA method for fetal heart rate extraction one maternal abdominal ECG and one fetal ECG extracted from JCA respectively From the fetal ECG you can see that the maternal ECG has been suppressed effectively and the FHR can be obtained accurately and conveniently ECG Baseline Remove One common variety of exercise induced artifacts is baseline wander resulting from movement respiration and poor electrode contact Although filters can be designed to remove much of this baseline variation they will distort the low frequency components of the ECG complex such as the TP segment the PR segment and most problematically the ST segment The
4. F Fig 16 Block diagram for designing and using a highpass filter to remove baseline wandering sampling frequency of 200Hz and 1000Hz Figure 15a and 15b shows an example of the specifications of the highpass filter and the block diagram of a sample virtual instrument that you can use to remove the baseline wandering eae Concerning the stability of the system the position of poles is inside the circle and the zeroes are inside and outside It is possible to observe in fig 15 a that zeros are equidistant to the unit circle for sampling frequency 200Hz and in fig 15 b for Sampling frequency 1000Hz It is as well possible to change the window For example window Dolph Chebyshev assures as well equidistance between zeros and the unit circle and the poles inside the circle The stability is assured as well in this situation ICA Independent Component Analysis is a method for retrieving independent hidden signals from a multi channel observation Assume that the observation X is a superposition of the source signal S it is possible to write X AS where A denotes a mixing matrix The assumption of statistical independence of the signals S allows you to estimate both unknowns S and A from the INTERNATIONAL JOURNAL OF CIRCUITS SYSTEMS AND SIGNAL PROCESSING observation X Here the maternal ECG and the fetal ECG can be treated as independent components i e they are parts of S while the 8 channel measured ECG records constitute X
5. ST segment is the most diagnostically relevant measure of the ECG taken during exercise While linear baseline interpolation and removal may be adequate at lower heart rates they also will introduce Significant distortions This is particularly evident when excessive nonlinear wander is present as seen at higher heart rates and respiration rates Since the cubic spline is not a filter and uses an a priori knowledge of the shape of the ECG signal it estimates the true baseline and avoids distortion better The more common implementations of this technique use relatively short ECG recordings With the advent of increasing power in computerized ECG systems the implementation of the cubic spline algorithm for removing baseline wander in continuous longer duration ECG records and in real time processing is being attempted However the correct application of the cubic spline to continuous recordings is not straightforward and involves a number of previously unforeseen difficulties The accuracy and resolution of both floating point and integer operations is Issue 4 Volume 1 2007 critical during long term application of the cubic spline function After removing the baseline wandering the resulting ECG signal is more stationary and explicit than the original signal However some other types of noise might still affect feature Abdominal ECG 0 6 koa In i 7 7 5 Amplitud Time s Fetal ECG eld l l 3 3 5 4 4 5 5 5 5 6 6
6. TXT data file is a file that contains only a 1D real array a 2Xm real matrix or an mx2 real matrix The virtual instrument processes the 1D real array from TXT data files as a real data type The virtual instrument processes the 2xm real matrix or the mx2 real matrix from a TXT data file as a complex data type The virtual instrument treats the first row or column as the real part of the data and treats the second row or column as the imaginary part of the data File Path Configuration contains the following options File path Specifies and displays the full path to the file from which we want to read data Ask user to choose a file each time this VI runs Displays a dialog box that prompts users to select a file each time the virtual instrument runs This Express Virtual Instrument displays the original signal and the signal after denoising The transform settings contain the following options Transform type Specifies the use of the discrete wavelet transform or undecimated wavelet transform to denoise the signal Wavelet specifies the wavelet type to use for the discrete wavelet analysis The default is db02 The options include two types orthogonal Haar Daubechies dbxx Coiflets coifx Symmlets symx and biorthogonal FBI Biorthogonal biorx_x where x indicates the order of the wavelet The higher the order the smoother the wavelet The orthogonal wavelets are not redundant and are suitable for signal or image deno
7. abdominal ECG and can be used for the extraction of fetal heart rate which indicates the cardiac condition of the fetus The locations of leads for an 8 channel maternal ECG acquisition system are shown in Figure 11 Maternal thorax ECG signals are sampled from thorax leads while maternal abdominal ECG signals are obtained from abdominal leads With LabVIEW and related toolkits such as the Advanced Signal Processing Toolkit ASPT and the Digital Filter Design Toolkit DFDT are shown in Figure 12 and Express VI Development Toolkit it is conveniently to build signal Thorax Leads Abdominal Leads Fig 11 Thorax leads and abdominal leads for maternal and abdominal ECG acquisition Issue 4 Volume 1 2007 322 Addons Time Frequen a ime Series A wavelet Anal Fig 12 Digital filter design and advanced signal processing processing applications for both stages including baseline wandering removing noise cancellation QRS complexes detection and fetal heart rate extraction In Fig 13 an ECG signal recorded from the abdomen of a pregnant woman is shown Simultaneously is recorded an ECG from the woman s chest M aking a comparison between the two signals we observe that the abdominal ECG demonstrates multiple peaks that mean QRS complexes corresponding to the maternal ECG occurring at the same time instants as the QRS complexes in the chest lead as well as others at weaker levels and a higher repeti
8. an efficient and cost effective method of patient biomedical signal acquisition and monitoring Fig 1 Computer based biomedical signal acquisition and analysis system M anuscript received A pril 11 2007 Revised version received November 18 2007 Mihaela Lascu is with Politehnica University Timisoara Faculty of Electronics and Telecommunicatons Department of Measurements and Optical Electronics Romania corresponding author to provide phone 0040 256 275488 e mail mihaela lascu etc upt ro Dan Lascu is with Politehnica University Timisoara Faculty of Electronics and Telecommunicatons Department of Applied Electronics Romania corresponding author to provide phone 0040 256 275488 email dan lascu etc upt ro Issue 4 Volume 1 2007 317 A computer based system consists of a few external hardware components for isolation and amplification of the signals a data acquisition card and a software analysis package as shown in Fig 1 Acquisition of the signal can be handled through built in procedures and LabVIEW s ability to easily create a user interface The analysis of the received signal can be performed by ready made procedures which can be obtained from N ational Instruments Il DATA ACQUISITION A wide range of data acquisition cards is available from National Instruments These cards provide for multiple channels of analog data input as well as output In addition LabVIEW comes with ready made l
9. maintains a consistent software interface among its different versions so that you can change platforms with minimal modifications to your code Whether we are using conventional programming languages or N ational Instruments application software our application uses the NI DAQ driver INTERNATIONAL JOURNAL OF CIRCUITS software as illustrated in Fig 2 Component Vorks LabVIEW LabWindows C y 1 or VirtualBench Conventional Programming Environment NLDAG Driver Software Personal Computer or Workstation DAG or STX Hardware Fig 2 Relationship between the programming environment NI DAQ and DAQ Card After installing NI DAQ device drivers we obtain on the desktop of the computer an icon named MAX M easurement and Automation Explorer This program is necessary for identifying the presence of the DA Q card and for settling the channels that will be used M easurement and Automation Explorer MAX provides access to our PCI 6023E DAQ acquisition card MAX can manage devices and interfaces the installed National Instruments software virtual channels or tasks for the used devices and as well it can create scales for the realised virtual instrument can configure the IVI Interchangeable V irtual Instrument drivers and import export the device configuration file An overview of the hardware functions of PCI 6023E board iS presented in Fig 3 Tiger l imig input l ai filet Timrgitenvrel Fossiibal DAQ STC maiwa
10. CESSING that passed the ECG signal could be subjected to a peak searching algorithm to obtain a time marker for each QRS or ECG beat The intervals between two such consecutive markers give the RR interval which could be averaged over a number of beats to obtain a good estimate of the inter beat interval The heart rate may be computed in bpm as 60 divided by the average RR interval in seconds The heart rate may also be obtained by counting the number of beats detected over a certain period IV LABVIEW FOR ECG MATERNAL AND FETAL SIGNAL PROCESSING Biomedical signal processing algorithms form an important part of real time systems for monitoring of patients who suffer from a life threatening condition Such systems are usually designed to detect changes in cardiac or neurological function and to predict the outcome of a patient admitted to the intensive care unit 1 2 Since such changes may be reversible with early intervention irreversible damage can sometimes be prevented Similar to therapeutic contexts the signal is processed during monitoring in an essentially sequential fashion such that past samples constitute the main basis for a decision while just a few seconds of the future Samples may also be considered a property which usually stands in sharp contrast to signal processing for diagnostic purposes where the signal is acquired in its entirety prior to analysis The fetal electrocardiogram can be derived from the maternal
11. Compatibility ACER Dan Florentin Lascu was born in Timi oara Romania on June 30 1961 He received the M Sc degree in electrical engineering and the PhD degree in electronics from Politehnica University Timi oara Romania in 1986 and 1998 respectively Since 1990 he has been with the Politehnica University Timisoara Applied Electronics Department and since 2007 is a Professor in the Power Electronics Group He published more than 65 papers and 3 books in the field of power electronics and his current research is the field of switching converter synthesis converter modelling converter simulation active power factor correction and soft switching techniques Dr D Lascu is a member of the Romanian Electronics Engineer Association AIE and a member of the Romanian Association of Electromagnetic Compatibility ACER
12. INTERNATIONAL JOURNAL OF CIRCUITS SYSTEMS AND SIGNAL PROCESSING Graphical Programming based Biomedical Signal Acquisition and Processing M ihaela Lascu Dan Lascu Abstract This paper describes a computer based signal acquisition processing and analysis system using LabVIEW a graphical programming language for engineering applications Biomedical signal acquisition has greatly advanced over the years using many different technologies E series multifunction data acquisition cards are used for the acquisition of biomedical signals and the appropriate software NI DAQ National Instruments Data Acquisition With the increasing performance of the personal computer computer based signal processing systems are becoming an efficient and cost effective way of acquiring and analyzing these signals The advanced analysis techniques available on the computer are becoming invaluable to the practicing physician The diagnostic decision will be more accurate Peak detection in electrocardiogram ECG is one of the solved problems using LabVIEW and filtering biomedical signals in different ways is a challenge that has to be solved Keywords biomedical signal data acquisition graphical programming language signal processing INTRODUCTION OMPUTERS are becoming a necessity in the medical community Physicians use computers for patient records and information It is obvious that personal computers based signal acquisition and analysis is
13. e introduced with LabVIEW that means the values and the name of the channel are selected from the controls belonging to the front panel the terminal sources from the diagram block take these new values and channel name and actualize the acquisition After running the above program we can indicate the three graphs that meet the necessity to publish the input signal the low pass filtered signal and the averaging with the Hanning filter The signals that have been acquisitioned and processed as in the above presented theory are visualised in following figures Fig 6 and Fig 7 Two acquisitions for two different electrocardiograms are done with the PCI 6023E card The easiest way to acquire a single waveform from a single channel is to use the Al Acquire W aveform virtual instrument 1 Using this virtual instrument it is necessary to specify a device and or channel the number of samples to acquire from the channel and the sample rate measured in samples per second It is also possible to set programmatically the gain by INTERNATIONAL JOURNAL OF CIRCUITS SYSTEMS AND SIGNAL PROCESSING setting the high limit and the low limit Using only the minimal set of inputs makes programming the virtual instrument easier but the virtual instrument lacks more advanced capabilities such as triggering Also itis possible to acquire more than one waveform at a time with another of the Easy Analog Input VIs Al Acquire Waveforms This virtual instrument ha
14. e uses one analog input line which connects to the positive input of the PGLA The negative input of the PGLA connects to analog input sense AISENSE In our application we use the RSE reference single ended configuration for example between AICH1 PIN 33 and AIGND PIN 32 for one channel The boards have a bipolar input range that changes with the programmed gain Each channel may be programmed with a unique gain of 0 5 1 0 10 or 100 to maximize the 12 bit analog to digital converter Ls BT mms a pio Ce Beli Bebe ho mj STi Pio pm aH Y aiam I ez pio Ce nim a CLT FES CAY CC oe le ah on ma ae Coa m ate ar pare m PT Eel PTT Ea m a m a a CaS a PAP osATo PPR TEn CoG a PPR CT Pel ZATE OPCIA oh FREO_ CAT J a ti EEGERERRBEE is CETATA OOE CGHO PrE DOO CET PrO ROTH ADURO PPG POTEH _ GATE GIT Ra_ OUT GoHo PPS AAT acl PPR PC Ti_ SOURCE DGHO TGHO lhis mailade on a Fig 4 1 0 connector pin assignment for the 6023E INTERNATIONAL JOURNAL OF CIRCUITS SYSTEMS AND SIGNAL PROCESSING ADC resolution With the proper gain setting you can use the full resolution of the ADC to measure the input signal Because it was not possible to make an acquisition of a biomedical signal in real time we have studied at the beginning the possibility for an electrocardiograph realisation that means the acquisition of a sinusoidal signal with an amplitude of 1 mV and
15. ed for each signal LabVIEW offers sweep scope and strip charts a Sweep chart sends a vertical line along with the signal from left to right so that the new data is to the left of the line while the old is on the right with a scope chart the signal travels from left to right and when it reaches the right edge the chart Issue 4 Volume 1 2007 321 0 69 0 87 3164 5 3164 5 lt threshp threshy 250 samples of EKG data decimated B from 1000 to SOsamples second B Maximum number of samples is 250 H aters IH actcrsr_ Fig 10 Peak detection in an electrocardiogram instrument named Peak Detector we can obtain the desired diagnostic decision These prebuilt modules are extremely useful in complicated applications We have established the peaks for different kinds of biomedical signals Peak detection is the first step in event detection 2 4 5 Finally we can say that event detection is an important step that is acquired before attempting to analyze the corresponding waves or wavelets After a specific wave of interest has been detected isolated and extracted methods targeted to the expected characteristics of the event may be applied for directed analysis of the corresponding physiological or pathological event We can use this Peak Detection in an application like an ECG Rhythm analysis The output of the final smoothing filter INTERNATIONAL JOURNAL OF CIRCUITS SYSTEMS AND SIGNAL PRO
16. f Amga l RT Bas Cage Tengt intasace Aani ipul pi Conii l iawa i Suri G i Tirang LAS V Connector PGI Connector for PCl 602 PAI Connector for PXI 6025E Fig 3 Typical block diagram of 6023E 6024E and 6025E acquisition cards Issue 4 Volume 1 2007 318 SYSTEMS AND SIGNAL PROCESSING The boards have three different input modes nonreferenced single ended NRSE input referenced single ended RSE input and differential DIFF input The single ended input configurations provide up to 16 channels The DIFF input configuration provides up to eight channels Input modes are programmed on a per channel basis for multimode scanning It is possible to configure the circuitry to scan 12 channels four differentially configured channels and eight single ended channels Table 1 describes the three input configurations If conference please contact your conference editor concerning acceptable word processor formats for your particular conference Table A channel configured in DIFF mode uses two analog input lines One line connects to the positive input of the board s programmable gain instrumentation amplifier PGIA and the other connects to the negative input of the POIA A channel configured in RSE mode uses one analog input line which connects to the positive input of the PGIA The negative input of the POIA is internally tied to analog input ground AIGND A channel configured in NASE mod
17. frequency of 1 2Hz from a function generator Then we have done the acquisition of an equivalent signal for a biomedical signal that is specific for the heart activity Using the following block diagram from Fig 5 we have an acquisition and a visualisation of the real signal and the processed signal Because there are a lot of imperfect connexions from an electronic point of view and there are a lot of long linking wires in a polluted electromagnetic waves environment and the input impedance is high the acquired signal is noise disturbed For filtering the disturbed signal it is necessary to cascade two filters the first filter is a low pass having a cutt off frequency of 40Hz and the second filter is an averaging Hanning filter 2 W hen an ensemble of several realizations of an event is not available synchronized averaging will not be possible We are then forced to consider temporal averaging for noise removal with the assumption that the processes involved are ergodic that is temporal statistics may be used instead of ensemble statistics As temporal statistics are computed using a few Samples of the signal along the time axis and the temporal window of samples is moved to obtain the output at various points of time such a filtering procedure is called a moving window averaging filter and is implemented in LabVIEW by help of shift registers Such a filtering procedure is called a moving window averaging filter in general the term moving
18. ibraries for interfacing with these DAQ cards Using these libraries programs for the data acquisition are quickly and easily made for allowing more time to be spent on the processing and analysis of the acquired signals We used National Instruments PCI 6023E board from the 6023E 6024E and 6025E family The 6025E features 16 channels eight differential of analog input two channels of analog output a 100 pin connector and 32 lines of digital I O The 6024E features 16 channels of analog input two channels of analog output a 68 pin connector and eight lines of digital 1 0 The 6023E is identical to the 6024E except that it does not have analog output channels The NI DAQ driver software is included with all National Instruments DAQ hardware NI DAQ is not packaged with SCXI Signal Conditioning Extended Instruments or accessory products except for the SCX1 1200 NI DAQ has an extensive library of functions that you can call from your application programming environment These functions include routines for analog input A D conversion buffered data acquisition high speed A D conversion analog output D A conversion waveform generation timed D A conversion digital 1 0 counter timer operations SCX self calibration messaging and acquiring data to extended memory NI DAQ also internally addresses many of the complex issues between the computer and the DAQ hardware such as programming interrupts and DMA controllers NI DAQ
19. ising and compression The biorthogonal wavelets usually have the linear phase property and are Suitable for signal or image feature extraction Threshold settings contain the following options Soft threshold specifies to use the soft thresholding method The default is to use the soft thresholding method If you remove the checkmark from the Soft threshold checkbox the virtual instrument uses the hard thresholding method Thresholding rule specifies the threshold selection rules The default is SURE which indicates that the virtual instrument uses the principle of Stein s Unbiased Risk Estimate SURE to estimate the threshold If you select Hybrid the virtual instrument finds a compromise between the SURE method and the Universal Method When the signal to noise ratio of the noisy signal is very low the virtual instrument uses the Universal Method to estimate the threshold If you select Universal the virtual instrument sets the threshold to sart 2xlog Ls where Ls is the signal length If you select Minimax the virtual instrument uses the Minimax Principle to estimate the threshold Rescaling method Specifies the method to use to estimate the noise variance at each level The default is single level which indicates that the noise is white and that the virtual instrument estimates the noise standard deviation from the wavelet coefficients at the first level The virtual instrument uses the noise variance to rescale the thresh
20. ition analysis systems These systems can be an inexpensive replacement for the costly stand alone signal specific systems currently in use The components necessary for a LabVIEW based acquisition and analysis system are inexpensive and readily available The experience necessary to program this type of system in LabVIEW is small and the numbers of libraries available from National Instruments is growing These developments show that a computer based system using LabVIEW can be an efficient alternative to stand alone equipment and as the speed and reliability of the computer increases there will be more and more of these systems available REFERENCES J Olansen E Rosow Virtual Bio Instrumentation Biomedical Clinical and Healthcare applications in LabVIEW Publishing House Prentice Hall PTR 2002 R M Rangayyan Biomedical Signal Analysis W iley ntersciene J ohn Wiley amp SONS INC 2002 A Aldroubi M Unser Wavelets in Medicine and Biology Publishing House CRC Press 1996 M Unser A Aldroubi and M Eden On the asymptotic convergence of B spline wavelets to Gabor functions IEEE Trans Inform Theory 38 2 864 872 1992 C Li C Zheng C Tai Detection of ECG characteristic points using wavelet transforms IEEE Trans on Biomed Eng pp 21 28 1995 1 2 3 4 5 INTERNATIONAL JOURNAL OF CIRCUITS SYSTEMS AND SIGNAL PROCESSING M Unser A Aldroubi and M Eden A family
21. m their internal implementation The block diagram is represented in Fig 9 Following steps are implemented simulation of the screen is refreshed The most common chart type for biomedical signals however is the strip chart strip charts bring the new data in on the right side of the screen while the old travels off the left One of the greatest benefits of this type of chart is the ability to view past data using a scrollbar attached to the chart Analysis of the biomedical signals can also be easily done in LabVIEW The graphical nature of LabVIEW allows even the beginning programmer to easily write programs to analyze data without having to worry about the syntax problems associated with most programming languages In Fig 10 is represented a peak detection in an electrocardiogram Front Panel and Block Diagram Using LabVIEW and the virtual Signal Source length Fl EKG 100 Peaks Found B Valleys Found B Peak Locations P lbs biomedical signal using a generation of a DC signal with seed valey Locations BANi filter order uniform white noise the second step is filtering the signal 1 P Peak Amplitudes BAM A5E valey Amplitudes S60 EEOERO with a bandpass filter the third step consists of two parts mia Peak 2nd Derv BBM FA35EF0 Valley 2nd verv SGT E updating the W aveform Graph cursors to represent the current values of the upper and lower cut off frequency models and checking to see if the upper o
22. o o Poles lt 0 60 0 40 Design Feedback m Filter Order Error Message 0 20 a 0 50 0 00 0 50 1 00 1 22 z OK Help Fig 15 a Front panel for designing and using a highpass filter to remove baseline wandering with sampling frequency 200Hz Cancel Issue 4 Volume 1 2007 E passband Main Settin 4 i IV Magnitude in dB stopband HEN Filter Type IHighpass 7 Filter Specification me Sampling foo 4 10 Frequency Hz xen 7 2 Passband Edge is a a a0 Frequency Hz 7 has z 30 Passband Ripple fi dB ane Stopband Edge p Frequency Hz 2 9 T Eis A a TE EE 0 0 1 0 20 3 0 40 50 60 7 0 80 90 10 7 Stopband z Frequency Hz Attenuation o w 1 25 7 Plane E 3 Unit Circle n 1 00 g Zeroes S Design Method Kaiser Window i o be ka 0 80 9 re 2 Poles o qx R m Design Feedback 0 60 o o a0 a eE Filter Order 74 4 o B Error Message 0 20 x g ea o 0 00 Q 0 11 l l Tt 1 31 1 00 0 50 0 00 0 50 1 00 1 22 OK Cancel Help 323 Fig 15 b Front panel for designing and using a highpass filter to remove baseline wandering with sampling frequency 1000Hz aseline wander removed by Kaiser Window FIR Filter pov Filter E DFD Filtering vi Original ECG signal SSS Waveform Chan NSarnip ble nin F Kaiser Window Highpass Filter filter out
23. of polynomial spline wavelet transforms Signal Process 30 2 141 162 1993 nae National Instruments LabVIEW User Manual NI Corporate H eadquarters 2000 xxx National Instruments LabVIEW Measurements Manual Corporate H eadquarters 2000 6 7 8 NI Mihaela Ruxandra Lascu was born in Timisoara Romania on M ay 7 1962 She received the degree in electrical engineering from the Politehnica University of Timi oara in 1986 and the Ph D degree in electromagnetic compatibility and measurement techniques in 1998 Since 1990 she was appointed Assistant at the University of Timi oara Her main interest is in numerical techniques applied in electromagnetic compatibility such as finite element analysis and finite difference time domain methods and their application to interference problems in steady state and time domain applications As well her main interest is in virtual instrumentation graphical programming and biomedical signal processing Now she is associate professor atthe University of Timi oara She published more than 60 papers and 2 books in the field of virtual instrumentation and her current research is the field of graphical programming biomedical signal processing and numerical techniques applied in electromagnetic compatibility Issue 4 Volume 1 2007 326 Dr M Lascu is a member of the Romanian Electronics Engineer Association AIE and a member of the Romanian Association of Electromagnetic
24. old Therefore you can update the thresholds with the Issue 4 Volume 1 2007 325 noise variance Selecting one indicates that the noise is white with unit variance Selecting multiple levels indicates that the noise does not have to be white and that the virtual instrument estimates the noise standard deviation at each level independently Option for approximation specifies the Operation for the approximation coefficients from the wavelet decomposition The default is none which indicates that the virtual instrument keeps the approximation coefficients unchanged If you select threshold the virtual instrument applies the same thresholding operation to the approximation coefficients If you select detrend the virtual instrument sets the approximation coefficients to 0 In the last few years 7 8 wavelets have emerged as a powerful tool for extracting signals from noisy data For the case of Gaussian white noise the signal extraction problem can be stated as follows we determine the true values of a signal S given set noisy observations S S 0 Z 1 0 n 4 where S S t and S S t at times t i n o is the standard deviation of the noise and z are random variables according to N 0 1 In the wavelet domainy we can rewrite WS W S 05 W S o W z 5 V CONCLUSION As the performance to cost ratios for computers continues to grow there will be a great need for computer based acquis
25. r lower cut off frequency values Siga E Peaks MEA valleys MA have changed since the last iteration if so it is necessary to reset the Dual Spectral M easurement averaging the fourth step performs a Dual Channel Spectral M easurement on the prefiltered signal and the filtered signal to determine the frequency response of the filter the last step test determines the calculated filter frequency response against preset specifications and determine whether the filter meets the requirements or not In the simulation we have used the uniform white noise because It represents the entire frequency range In this way the Dual Channel Spectral M easurement ia H amp i Dual Channel Spectral i Measurement p Signals Input Signal Tested Signals Input SignalB Passed Magnitude Y Point Evaluation Simulated Signal Heh Simulate Signal 4 Mask and Limit Waveform Graph Testing Passed Filtered Signal gt Filtered Signal hap EL 7 H atas Cursor Posx TH atc Cursor Posx Upper Cut Off Filter Frequency Response Magnitude mz Lower Cut Off D 4 Fig 9 A dual channel spectral biomedical signal measurement with filter using Express V Is Express virtual instrument easily determine what frequencies are being filtered out and by how much For the signal display several signals can be displayed on the same chart or separate charts can be configur
26. s also a minimal set of inputs but it allows inputs of more than one channel to read and return an array of waveforms from all channels it reads To access or control an individual waveform index the array of waveforms with the Index Array function or use input indexing on a For or While Loop A two channel data acquisition system can be realized iri i Yi i i h l r i h l i ty I F Wy iN y TAg un Wy Fig 6 Electrocardiogram and low pass filtered electrocardiogram Semnal cu zgomot de B0H z Ampltudines Fig 7 Electrocardiogram with 60H z noise first graph filtered electrocardiogram second graph and averaged electrocardiogram Hanning filter third graph as in Fig 8 Issue 4 Volume 1 2007 320 Channel 1 Data Channel 2 to 00 00 00 DOMM YYYY dt Channel 1 spri Channel 1 Channel 2 Frequenc Frequenc 200 200 0 000000 150 cae 100 a cp 50 Channel 2 Data 00 00 00 sampling rate DEMME YYY LOU OL dt 0 000000 points j 1024 Fig 8 Two channel data acquisition system We have in this program the possibility to choose the channel the frequency the device and the sampling rate To acquire multiple waveforms it is also possible using the Intermediate virtual instruments IIl SIGNAL PROCESSING COMPUTER BASED ANALYSIS AND DISPLAY The biomedical signals acquired from the human body are frequently
27. sing Wideband Noises Suppression Baseline Wandering Removing Features Extraction QRS Complexes Extraction Fetal ECG Extraction Suppresses noise from the raw ECG signal and the feature extraction stage extracts diagnostic information from the ECG signal Preprocessing ECG signals is necessary to remove contaminants from the ECG signals Broadly speaking ECG contaminants can be classified into the following categories power line interference electrode pop or contact noise patient electrode motion artifacts electromyography EMG noise and baseline wandering The LabVIEW provides an intuitive and interactive way to design and implement finite impulse response FIR or infinite impulse response IIR filters easily and effectively We used the Configure Classical Filter Design Express VI to design a Kaiser Window FIR highpass filter to remove the baseline wandering The configuration is realized for a E passband E stopband HAW m Main Settin i IV Magnitude in dB IHighpass Filter Type m Filter Specification Sampling Frequency Hz Passband Edge Frequency Hz Passband Ripple 60 70 l l l I l l l 00 1 0 20 3 0 40 50 60 7 0 80 9 0 Frequency Hz Unit Circle 4 Zeroes Stopband Edge Frequency Hz l T 5 J Stopband Attenuation 20 a dB Kaiser Window 1 26 o P o 2 Plane 1 00 0 80 o o Q gt o AS x Design Method FE
28. tion rate The ha r p taf wy Meena May aj Weta a N 1 l j i Cal r WOTE IER FETUS l al j Fa il J i I Li pot J nel Fo A j d i r th Fig 13 ECG signals of a pregnant woman from abdominal and chest leads a chest leads ECG and b abdominal lead ECG non maternal QRS complexes represent the ECG of the fetus The former presents the maternal ECG whereas the latter is a combination of the maternal and fetal ECG signals It is necessary to observe that the QRS complex shapes of the maternal ECG from the chest and abdominal leads have different shapes due to the projection of the cardiac electrical vector onto different axes Considering that the two signals being combined have almost the same bandwidth it is necessary to separate them and obtain the fetal ECG Generally the recorded ECG signal is often contaminated by noise and artifacts that can be within the frequency band of interest and manifest with similar characteristics as the ECG signal itself In order to extract useful information from the noisy ECG signals it is necessary to process the raw ECG signals INTERNATIONAL JOURNAL OF CIRCUITS SYSTEMS AND SIGNAL PROCESSING ECG signal processing can be roughly divided into two stages by functionality preprocessing and feature extraction as shown in Fig 14 The preprocessing stage removes or on lt Fig 14 Preprocessing ECGs records and feature extraction Preproces
29. very small often in the millivolt range and each has its own processing needs Electroencephalography signals are in the microvolt range and have many frequency components Obviously these biomedical signals require processing before they can be analyzed LabVIEW contains the tools from fast Fourier transforms to digital filters to realize complex analysis In order to do frequency analysis a complex signal must first be broken down into its frequency components One of the most common way to do this is with an FFT In order to facilitate this type of analysis LabVIEW comes with built in FFTs that make the process of component INTERNATIONAL JOURNAL OF CIRCUITS SYSTEMS AND SIGNAL PROCESSING separation quick and easy In addition biomedical signals being very small are in danger to be overwhelmed by noise TO combat this it is necessary to use a SCXI card that means to run the acquired signal through a set of filters and amplifiers 5 However after the signal reaches the computer it can still contain noise Another way to solve the noise problem is to use the digital filters provided with LabVIEW LabVIEW offers the choice of Butterworth Bessel Chebyshev and digital filters With a few adjustments these filters can be configured for almost any design that is needed A dual channel spectral measurement with filter is presented using the Express virtual instruments that have the possibilities to choose different kind of parameters fro
Download Pdf Manuals
Related Search