Home

Acknoweldgments - Introduction

1. ADS535 16 BIT Figure 7 1 Data path flow within the DSP board Amplified ECG ECG with pacing markers The ISR for the serial port receive interrupt Appendix 3 source file dss_cisr copies each new 32 bit data sample in the Data Receive Register DRR Appendix 3 source file dss_dsk6211 c to a frame ofthe DSS_rxPipe Appendix 3 source file dss c When the frame is full the program puts the frame back into DSS_rxPipe to be read by the audio function Appendix 3 source file audio c As the audio function will just read a frame and copy it the transmit rate will be the same as the receive rate 48 kHz That is why only the receive interrupt for the serial port was enabled The transmit interrupt for the serial port is not enabled The program was later changed to include the required extra processing to fulfil the objectives of the project It was decided to continue with the above principle as it did not cause any delays or distortion to the output data The 48 kHz sampling rate employed here is not the rate at which real world pacemakers are used That is because of the limited data area higher sampling rate means more samples to store and restrictions on power consumption that implantable devices have to operate The quantisation of the ADC IS 16 bit and cannot be changed Epameinondas Petropoulos 21 MSc in Electronics A Programmable Pacemaker University of Hertfordshire The sample rate could be imitated to be lo
2. amplifier These connections are the outputs at the rear of the Heartsim 2000 The input signal to the amplifier consists of the desired biopotential power line interference of 60 Hz and its harmonics and noise In addition to these elements the Heartsim gives the chance to experience simulated muscle noise and even more power line interference Proper design of the amplifier rejected a large portion of these interferences The differential amplifier has a central role in our design The main task is to reject interference coupled into the signal The ECG appears as a voltage between the two input terminals differential signal Epameinondas Petropoulos 21 MSc in Electronics A Programmable Pacemaker University of Hertfordshire The line frequency interference signal causes approximately the same potential at both inputs common mode signal Strong rejection of the common mode signal is one of the most important characteristics of a good biopotential amplifier The common mode rejection ratio CMRR of an amplifier is defined as the ratio of the differential mode gain The CMRR of the TL071 op amp is 86 dB typical The rejection of the common mode signal in a biopotential amplifier is a function of both the amplifier CMRR and the source impedances Source impedances ideally should be equal If they are not the common mode voltage drop shows a differentialsignal at the amplifier input unwanted In practice equal resistors do not exist a
3. course information material pp 2 12 3 Webster Encyclopaedia chapter on CPR pp 600 620 4 NASPE Educational Guidelines Pacing and Electrophysiology 2nd Edition p 47 5 Mirowski M et al The automatic implantable cardioverter defibrillator J Am Coll Cardiol 1985 6 461 66 6 Davies DW et al Better recognition of arrythmias by implanted devices Proc 9th International Congress The New Frontiers of Arrhythmias 1990 pp 63 7 7 Jenkins J et al Impact of filtering upon ventricular tachycardia identification by correlation waveform analysis PACE 1991 14 661 8 Throne RD et al A comparison of four new time domain techniques for discriminating monomorphic ventricular tachycardia from sinus rhythm using ventricular waveform morphology IEEE Trans Biomed Eng 1991 38 gt 561 70 9 TLO71 Data sheet p 3 10 JD Bronzino The Biomedical Engineering Handbook p 1187 11 JD Bronzino The Biomedical Engineering Handbook p 1298 12 http www s ti com sc psheets spra598 spra598 pdf 13 NASPE Educational Guidelines Pacing and Electrophysiology 2nd Edition p 12 14 SG Kochan Programming in C p 147 15 JG Webster editor Encyclopaedia of Medical Devices and Instrumentation Vol 1 p 604 6 16 NASPE Educational Guidelines Pacing and Electrophysiology 2nd Edition p 33 17 JG Webster editor Encyclopaedia of Medical Devices and Instrumentation Vol 1 p 2176 Epameinondas Petropoulos 54 MSc in Electronics A Programma
4. enough Overall DSP s are not needed for pacemaker only algorithms Epameinondas Petropoulos 20 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Chapter 6 Biopotential Amplifier The purpose of this chapter is to introduce the amplifier which was designed and built for amplifying the ECG signal Why it was needed what are its requirements how they were met schematic diagrams and output graphs are all to be found in this chapter 6 1 General Specifications The ECG signal represents the sum of the electrical potentials of all the depolarising tissue of the heart Real practice ECG measurements and the output of the Heartsim 2000 involve voltages around 2mV with high source impedance and as an option on Heartsim 2000 superimposed high level interference signals and noise The signal needs to be amplified to utilise the full range of the A D converter which is included in the DSP board The amplifier must be adequate to measure the signal It has to provide amplification selective to the ECG signal and reject everything else 6 2 Basic Requirements Two basic requirements apply in our case The amplifier should not distort the measured signal The amplifier should provide the best possible separation of signal and interferences Three electrodes two of them picking up the biologic signal RA LA and the third RL providing the reference potential make up the connection of the human subject to the
5. 1 Definition states that P R intervals are actually prolonged until conduction fails and then the node recovers A 4 3 conduction ratio means 4 P waves to 3 QRS complexes Figure 8 11 shows a representative waveform of 2 QRS Type E The emphasis isnot on the QRS type but on the conduction ratio j d A m r s m Figure 8 11 4 3 conduction ratio of 2 AV block Type 1 Source Heartsim 2000 User Manual The result expected was 1 pacing beat for every 3 naturals Figure 8 12 shows the result The pacemaker is functioning as it should It sensesthe natural beats correctly Then it fires a pacing pulse when a natural pulse is missing the blank interval is larger than the average of the three previous intervals plus 10 b Figure 8 12 4 3 conduction ratio of 2 AV block Type 1 Pacing pulses appear positive and wide a QRS Type E Same type as Heartsim Manual example of Figure 8 11 Natural beats appear biphasic fifth appears positive b QRS Type A Natural beats appear inverted Epameinondas Petropoulos 43 MSc in Electronics A Programmable Pacemaker University of Hertfordshire 2 2 AV block Mobitz type 2 3 1 2 1 Definition states that some atrial impulses will fail to conduct a ventricular contraction Some P waves are not conducted A 3 1 conduction ratio means that there are 2 non conducted P waves to 1 conducted ORS A 2 1 conduction ratio means that we have non conducted P wave
6. 2 2 8 2 4 and 8 2 5 Epameinondas Petropoulos 51 MSc in Electronics A Programmable Pacemaker University of Hertfordshire 9 3 Time Plan and Cost The time plan as was set in the feasibility report was not followed exactly There was considerable delay in getting started with the DSP board look Appendix 2 Problems with the DSP Three quarters of the time were spent troubleshooting the DSP During that period the algorithm was developed in paper When the problems with the DSP were solved the algorithm was implemented quickly and successfully The equipment used was not bought specifically for the project The only costs made are the 7 TLO71 op amps costing 0 30 each There is no commercial viability of the project as it stands now 9 4 Suggestions for future work and Improvements Having as basis the pacemaker presented in this report the most obvious way to go further is to incorporate with a cardioverter defibrillator Vhis is most interesting as there is still research going on unlike the pacemaker To incorporate a cardioverter defibrillator the system must be able to recognise arrhythmias Arrhythmia recognition is an evolving field and there are numerous methods to experiment with The C6211 is able to support the heavy computational demands that many algorithms command To proceed further it is assumed that one finds the system presented here satisfactory To test it under more realistic conditions another signal generator m
7. 8 3 Normal Sinus Rhythm rate 80 QRS Type A The columns expected and result refer to the state of the pacemaker Activated means it is pacing at regular intervals or when needed Inhibited means there is no pacing at all In most cases the estimate of what was expected was based on the rate only Table 8 2 QRS Type A in different rhythms and rates Rhythm Rate Expected Result Sinus Bradycardia 30 Activated Activated 40 Activated Activated 50 Activated Activated Normal Sinus 60 Inhibited Inhibited Rhythm NSR 80 Inhibited Inhibited Sinus Tachycardia 100 Inhibited Inhibited 120 Inhibited Inhibited 140 Inhibited Inhibited Atrial Tachycardia 90 Inhibited Inhibited 180 Inhibited Inhibited Epameinondas Petropoulos 35 MSc in Electronics A Programmable Pacemaker University of Hertfordshire 260 Inhibited Inhibited Atrial Flutter 75 Inhibited Inhibited 100 Inhibited Inhibited 150 Inhibited Inhibited Atrial Fibrillation 40 Activated Activated 90 Inhibited Inhibited 180 Inhibited Inhibited Junctional 30 Activated Activated 80 Inhibited Inhibited 180 Inhibited Inhibited Ventricular note 1 120 Inhibited Activated Tachycardia 180 Inhibited Activated 210 Inhibited Activated Ventricular note 2 Very Coarse Activated Activated Fibrillation Standard Activated Ac
8. apart from its QRS complex as only the ventricular activity is sensed by a VVI pacemaker According to the decision rules applied a pacing pulse fires or not The rule of thumb used was that the pacemaker should become active whenever the interval between QRS complexes becomes larger than sec less than 60 bpm Some tolerance was added to accommodate for normal heart variability The resulting pacing pulses were diplayed in the same chart with the QRS complexes The display consisted of a data acquisition card hosted in a Pentium PC and running on an in house made program called TCSCOPE As the project tries to emulate a VVI pacemaker only signals sensed from the ventricles are of interest An intracardiac electrogram interpretation depends on the catheter location A catheter placed onthe right ventricle show a V wave That corresponds to the QRS wave on the surface ECG By stripping out the ECG signal apart from its QRS complex an accurate representation of intracardiac ventricle signal is achieved Extraction of the QRS complex from the ECG signal was achieved with a high degree of consistency A general block diagram is shown to help you visualise the process Figure 1 1 Epameinondas Petropoulos 2 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Amplifier Bandpass Filter PC amp DISPLAY Figure 1 1 General Block Diagram 1 3 Guide to the Report Chapter 2 is a brief introduction to electrocardiogr
9. are happening just before the natural pulse That is because our variability measure is 10 and extends the waiting time Natural beats terminate pacing both in real pacemakers and here Epameinondas Petropoulos 39 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Figure 8 7 2nd degree block type 1 5 4 conduction ratio Natural beats appear biphasi and one appears negative due to sampling problems with the 711 data acquisition card All paced pulses appear positive and wider Vertical axis units are volts Horizontal axis units are seconds 4 2 AV heart block Mobitz type 2 3 1 Definition of 2 AV block Mobitz Type 2 states that some atrial impulses will fail to conduct a ventricular contraction Some P waves are not conducted A 3 1 conduction ratio means that we have 2 non conducted P waves to 1 conducted QRS In between these cycles we might have sequences of normally conducted pulses Figure 8 8 shows a representative waveform of QRS type B The emphasis is not on the QRS type but on the sequence of conducted or non conducted pulses ae eee Figure 8 8 3 1 conduction ratio of 2 AV block Type 2 Source Heartsim 2000 User Manual Figure 8 9 shows the result of applying to the pacemaker such a waveform It can be seen that for the two non conducted P waves the pacemaker takes control and replaces the natural beats Whenever there is a natural beat the pacemaker is inhibi
10. in Electronics A Programmable Pacemaker University of Hertfordshire memory organisation The contents of the starter kit used follow next and presentation proceeds by illustrating the tools available for debugging the software The chapter closes with a critical evaluation of the usefulness of the C6211 in relation with the requirements of the project In appendix 2 there is a very useful list for fellow students detailing all the problems encountered during the project concerning the DSP as well as some ways to overcome them are also included In chapter 6 the biopotential amplifier is presented Its general specifications and basic requirements are deployed in order to create the framework upon which the design choices were made The ECG signal characteristics are revisited A differential amplifier and a band pass filter together with some extra gain make up the design Avdetailed schematic diagram is explained Results before and after amplification are shown Chapter 7 outlines the development of software employed and written to perform the signal processing After a brief on the requirements of such a real time task the audio example utilised by the project is explained in detail A reference to the pipes used to transfer data and a comprehensive data path flow are presented The sampling rate question is discussed Presentation of the algorithm developed and embedded into the audio example follows The threshold detec
11. it can be seen from the execution graph Processes can be hardware interrupts software interrupts or periodic objects All processes can be set up graphically and relatively easy Another useful feature is the CPU load graph The execution graph and the CPU load graph are shown in figure 5 2 The only software interrupt in the program is the audioS WI The CPU with the full program up and running is loaded only a E Project r hh teCnt aa Eg audio_dsk mak if DSS_txCnt 0 audio_dskcfg c txDone 1 9 DSP BIOS Cor M H Include i E Libraries kas th E c Source DSS_error 0x2 te AUDIO C audio_dske Ki Dss c if rxDone txDone 0 as return rete D dss_asm st 2 dss_cisr c Ri if esa i don t have to set writerSize or writerAddr m PESEE E E E E E EA EEE ta Ks 1 audioS wi waiting No Errors No Warnings tren ihrearte O ready Ink6x audio_dsk mak PRD Ticks m unknown THS320C6x COFF Linker Time a error Copyright c 1996 1999 Assertions E unning Build Complete eens E done 0 Errors 1 Warnings 4 4 1l 2 Md Last 2 99 0 1 TD d DSP RUNNING For Help press F1 aa a Peak 25 202 Figure 5 2 The Code Composer Studio environment 5 4 Evaluation of C6211 in relation to the objectives of the project The C6211 is designed with telecom applications in mind such as ADSL multi channel modems and GSM vo
12. natural 30 AV bl ck 30 Activated Activated Epameinondas Petropoulos 41 MSc in Electronics A Programmable Pacemaker University of Hertfordshire 50 Activated Activated 60 Inhibited Inhibited 8 2 4 QRS Type E Type E QRS complex is a biphasic QRS with T inversion Table 8 5 QRS Type E in different rhythms and rates Rhythm Rate Expected Result Sinus Bradycardia 30 Activated Activated 40 Activated Activated 50 Activated Activated Normal Sinus 60 Inhibited Inhibited Rhythm 80 Inhibited Inhibited 100 Inhibited Inhibited S Tachycardia 120 Inhibited Inhibited A Tachycardia 180 Inhibited Inhibited A Fibrillation 90 Inhibited Inhibited Junctional 80 Inhibited Inhibited V Tachycardia 180 Activated Activated V Fibrillation Activated Activated lo AV block 40 Activated Activated 60 Inhibited Inhibited 85 Inhibited Inhibited 20 AV block 1 5 4 1 paced 4 naturals 1 paced 4 naturals 4 33 note 1 1 paced 3 naturals 1 paced 3 naturals 3 2 1 paced 2 naturals 1 paced 2 naturals 20 AV block 2 3 1 2 paced 1 natural 2 paced 1 natural 3 1 2 1 note 2 interchanged interchanged 2 1 1 paced 1 natural Activated 30 AV block 30 Activated Activated 50 Activated Activated 60 Inhibited Inhibited Epameinondas Petropoulos 42 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Notes 1 2 AV heart block Mobitz type
13. remote control keyboard It comes with an extensive ECG library There are 8 types of QRS complexes see Table 4 1 and 14 basic rhythms see Table 4 2 The operator can increase or decrease the pulse rate introduce premature complexes and add interference of house current 60 Hz or muscular artefacts The premature complex types available are Unifocal premature ventricular complexes Multifocal premature ventricular complexes Couplet premature ventricular complexes Premature atrial complexes Premature junctional nodal complexes For definitions of the basic rhythms and premature complexes please refer to Appendix 1 Table 4 1 Description of supraventricular QRS types available QRS Type Description for pictures look appendix X A Normal upright QRS T Upright QRS with ST depression and T inversion B C Upright QRS with ST elevation D Bundle Branch Block broad R wave with T inversion BiPhasic QRS with T inversion QS with ST elevation G Bundle Branch Block broad S wave with upright T wave Epameinondas Petropoulos 14 MSc in Electronics A Programmable Pacemaker Table 4 2 Basic rhythms and their rate range University of Hertfordshire Basic Rhythms Rate Normal Sinus Rhythm Sinus 30 170 Bradycardia Sinus Tachycardia Atrial Tachycardia 90 260 Atrial Fluttering 75 150 Atrial Fibrillation 40 180 Junctional Nodal Rhyt
14. 7 Software for Pacing 25 Epameinondas Petropoulos MSc in Electronics A Programmable Pacemaker University of Hertfordshire 7 1 Requirements of the System 25 7 2 Utilising an Example Program 25 7 3 The Pacing Algorithm 28 7 3 1 The Threshold Stimulation 28 7 3 2 Pacing Decision 29 7 3 3 Updating the Escape Interval 29 7 4 Shortcomings and evaluation of code 32 Chapter 8 Results 33 8 1 QRS Types 33 8 2 Basic Rhythms 35 8 2 1 QRS Type A 35 8 2 2 QRS Type B 38 8 2 3 QRS Type C 41 8 2 4 QRS Type E 42 8 2 5 QRS Type F 45 8 3 Premature Beats 47 8 3 1 Premature Atrial Complexes 47 8 3 2 Premature Nodal Junctional Complexes 48 8 3 3 Premature Vnetficular Complexes 48 8 4 Artefacts and Noise 49 Chapter 9 Conclusions and Discussion 51 9 1 Summary of Work and Results 51 9 2 Objectives andh w they were met 51 9 3 Time Plan and Cost 52 9 4 Suggestions for Future Work and Improvements 52 9 5 Conclusion 52 References 54 Bibliography 55 APPENDICES Appendix DDefinitions of Basic Heart Rhythms 56 Appe dix 2 Problems with the DSP 58 Appendix 3 C Source Files 59 Epameinondas Petropoulos MSc in Electronics A Programmable Pacemaker University of Hertfordshire Chapter 1 Introduction During our Medical Electronics degree we came across a wide range of diagnostic and therapeutic devices dealing with all aspects of human malfunction Fron all those devices and processes none is more exciting than the implantable devices especially those dealin
15. A Programmable Pacemaker University of Hertfordshire Acknoweldgments Ray Gordon my supervisor for his support and advice through the period of the project Steven Passmore lab technician for his help in setting up the equipment his interest in the project and for all his patience with my demands George Pissanidis research student for his help in dealing with problems with the DSP Epameinondas Petropoulos MSc in Electronics A Programmable Pacemaker University of Hertfordshire CONTENTS Acknowledgments Chapter Introduction 1 1 Aims and Objectives 1 1 2 Overview of the project 2 1 3 Guide of the report 3 Chapter 2 Electrocardiography 5 2 1 Cardiac Anatomy and Function 5 2 2 Biopotentials in the Heart 6 2 3 The Electrocardiogram 6 Chapter 3 Pacemakers 9 3 1 Indications for Pacing 9 3 1 1 Disorders of Impulse Conduction Conduction Delay and AV Block 9 3 1 2 Disorders of Impulse Formation Sick Sinus Syndrome 10 3 1 3 Ventricular Tachycardia 10 3 1 4 Ventricular Fibrillation 10 3 2 The Pacemaker 11 3 2 1 VVI Pacemakers 11 3 3 Modalities 12 Chapter 4 The Heart Simulator 14 Chapter 5 DSP Boardand Development Tools 16 5 1 Introducing the TMS320C6211 C6211 16 5 2 C6211 Contents 17 5 3 Code Composer Studio 18 5 4 Evaluation of C6211 in relation to the objectives of the project 19 Chapter 6 Biopotential Amplifier 21 6 1 General Specifications 21 6 2 Basic Requirements 21 6 3 The Amplifier 23 Chapter
16. Fibrillation 90 Inhibited Inhibited Junctional 80 Inhibited Inhibited V Tachycardia 180 Activated Activated V Fibrillation Activated Activated lo AV block 40 Activated Activated 60 Inhibited Inhibited 85 Inhibited Inhibited 20 AV block 1 5 4 1 paced 4 naturals 1 paced 4 naturals 4 3 1 paced 3 naturals 1 paced 3 naturals 3 2 note I 1 paced 2 natural Inhibited 20 AV block 2 3N 2 paced natural 2 paced 1 natural 3424 interchanged interchanged 2 1 note 2 1 paced 1 natural Activated 30 AV block 30 Activated Activated 50 Activated Activated 60 Inhibited Inhibited Epameinondas Petropoulos 45 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Notes 1 2 AV block Mobitz type 1 3 2 In this case there is a 3 2 conduction 3 P waves to 2 ORS complexes Figure 8 15 shows a representative waveform of QRS type F gt fi h Aa oo WO a Wn fe ae F Lo 4 Figure 8 15 2 block type 1 3 2 QRS Type F Source Heartsim 2000 User Manual It is expected that the pacemaker fires a pulse every two natural beats In figure 8 16 the results can be seen It does fire a pacing pulse for every two naturals but it fires very late Figure 8 16 2 degree block type 1 3 2 Naturah beats appear positive and narrower than pacing pulses Vertical axis units are volts Horizontal axis units are seconds 2 2 AV block Mobitz type 2 2 1 With a 2 1 conduction ratio itis expected that
17. TINA drawing package Figure 5 2 shows a graph of the ECG signal with no amplification as it was coming out of the Heartsim Figure 5 2 Heartsim output with no amplification Figure 5 3 shows an amplified version of the ECG signal from the amplifier of Figure 5 1 The only difference is that Figure 5 2 was amplified with a lower cut off frequency of 2 Hz and all the T wave energy has been preserved The effect of saturation is evident on the QRS wave Figur 5 3 Amplified ECG signal QRS Type A rate 80 Epameinondas Petropoulos 24 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Chapter 7 Software for Pacing Key Points Sample the ADC at Regular Intervals Peak Detection Measuring Interval in Between Peaks Respond to changes in the rate of change of heart rate Blind Time after QRS detection 7 1 Requirements of the system The pacemaker s function is essentially a hard real time task in that it must respond to cardiac events within strict deadlines Therefore such a system that has a very strict response time and must not be allowed under any circumstances to crash because the consequences are life threatening Cardiac events happen at relatively long intervals compared with other non medical applications that DSPs and software has being written for The system must continuously receive data from the heart and in our case it must also continuously transmit data s
18. all the time Pacemaker pulses properly timed can break up such a tachycardia However such treatment can also elicit fibrillation 3 1 4 Ventricular Fibrillation Fibrillation is the uncoordinated and uncontrolled beating of small muscle groups of the heart each trying to pump blood on its own The heart pumps no blood The condition is usually fatalunless promptly treated with a defibrillator Cardioverter defibrillator functions can be incorporated into a pacemaker More usually it is vice versa Epameinondas Petropoulos 10 MSc in Electronics A Programmable Pacemaker University of Hertfordshire 3 2 The pacemaker A pacemaker is an electronic device which stimulates the heart when it would otherwise beat too slowly It consists of a small package containing the power source and control electronics and a lead which conveys stimulation pulses to the heart and_monitoring signals from it The package is placed in the body outside the ribs The lead is pushed along an arm vein into the ventricle in the case of VVI pacemakers For the pacing pulse a voltage of around 5 V is required with a pulse duration of 0 5 msec 3 2 1 VVI pacemakers The first implantable pacemakers simply fired a 2 msec_pulse into the heart The trend from there onwards is towards controlling the pacemaker rate and function For patients who have near normal heart function and only occasionally experience a syncope a fixed rate pacemaker would compete against the
19. aphy and the importance of the ECG signal in diagnosis It covers a brief description of heart anatomy and function and then proceeds to define the five waves P Q R S and T that constitute the ECG signal and their correlation to heart activity A definition of what constitutes a healthy heart rate and its correlation to the cardiac output finishes the chapter Chapter 3 introduces the concepts behind pacemakers and their functions It starts with the indications of abnormalities of heart activity that lead to the prescription of pacemakers Atrio Ventricular blocks are explained in some detail among other indications and their ECG characteristics are highlighted The VVI pacemaker is defined and its principles of operation are explained in detail The chapter finishes with the description of the different modalities ofpacemakers as formulated by North American Society of Pacing and Electrophysiology NASPE association Chapter 4 is a small section introducing the Laerdal Heartsim 2000 the heart simulator The variety of different ECG waveforms are mentioned A more detailed definition of thebasic rhythms and premature complexes can be found in Appendix 1 Chapter five finishes the section of background material by introducing the DSP board C6211 and development tools that come with it It is a 150 MHz DSP from Texas Instruments A few things about its internal architecture are presented with bias towards its Epameinondas Petropoulos 3 MSc
20. ble Pacemaker University of Hertfordshire BIBLIOGRAPHY 1 JJ Carr JM Brown Introduction to Biomedical Equipment Technology 3rd edition 2 L Cromwell et al Biomedical Instrumentation and Measurements 2nd edition 3 R Aston Principles of Biomedical Instrumentation and Measurement 4 JG Webster editor Medical Instrumentation Application and Design 3rd edition 5 RS Khandpur Handbook of Biomedical Instrumentation 6 JG Webster Encyclopaedia of Medical Devices and Instrumentation Epameinondas Petropoulos 55 MSc in Electronics
21. ced 3 naturals 32 1 paced 2 naturals 1 paced 2 natural 20 AV block 2 3 note 2 2 paced 1 natural 2 paced 1 natural 342 interchanged interchanged 2a 1 paced 1 natural Activated 30 AV block 30 Activated Activated 50 Activated Activated 60 Inhibited Inhibited Epameinondas Petropoulos 38 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Notes 1 2 AV heart block Mobitz type 1 5 4 Definition for Mobitz type 1 2 AV block states that P R intervals are actually prolonged until conduction fails and then the node recovers A 5 4 conduction ratio means 5 P waves to 4 QRS complexes Figure 8 6 shows a representative waveform of QRS type D The emphasis is on the sequence of conducted and non conducted QRSs 1 f i if Ee ea P O Figure 8 6 2nd degree block type 1 5 4 conduction ratio Source Heartsim 2000 User Manual Expected result is 1 paced pulse for 4 naturals Figure 8 7 shows the result It can be observed that the second and the last pacing pulses are occurring as expected The interval between the first and second natural beats is 2 2 sec This is where the P wave was not conducted The interval between the third and the second heartbeat is 1 7 sec 35 bpm and pacing is justified The interval between the fourth and third natural beats is 1 6 sec pacing again justified The interval between fifth and fourth is again 1 6sec and pacing is justified again Pacing pulses
22. coders Its advantage is that it can perform high speed number crunching routines in real time while relying in inexpensive external memory for program and data storage For a pacemaker design external memory should be avoided if possible due to power consumption and extra bulk The designer should concentrate on utilising the 72 Kbytes of on chip memory of which only 4kbytes are available for program memory Assembly programming for greater efficiency is needed If monitoring capabilities are required the remaining 64 kbytes can be configured as RAM allowing electrogram records of Epameinondas Petropoulos 19 MSc in Electronics A Programmable Pacemaker University of Hertfordshire approximately 4 min duration which is about the state of the art today Actual duration will depend on sampling rate and compression techniques As for higher speed number crunching in real time it would only come of use if dealing with arrythmia recognition algorithms with high computing needs Such algorithms might be the time domain morphology analysis Probability Density Function PDF or the Gradient Pattern Detection GDP Correlation Waveform Analysis is also known to make heavy demands on computation These methods can be used with slower machines which might have very limited memory by being invoked only when there is a possibility of arrhythmia For simple pacemaker designs such a VVI design a general purpose microprocessor should be more than
23. conduction path has excessive delay so ventricular depolarisation is late This can cause faintness during exercise as the ventricles may not have time to fill properly during diastole but the condition often shows no symptoms Usually a pacemaker is not needed ECG characteristics Each P wave is conducted but the P R interval exceeds 0 2 sec and remains constant Rhythnvis regular and QRS is normal 2nd Degree Block The conduction path is bad enough to cause occasional signals to be too weak to fire the ventricles The ventricles will eventually fire spontaneously but the beat will be verylate and progress through the heart from apex to base causing an inverting ECG waveform If the late and slow beats occur frequently the condition causes temporary weakness and breathlessness and a pacemaker may be prescribed ECGrcharacteristics At regular or irregular intervals a P wave is not followed by a QRS complex The P P interval is constant Two types of 2 degree block can be found Epameinondas Petropoulos 9 MSc in Electronics A Programmable Pacemaker University of Hertfordshire The less serious Mobitz type I where succeeding P R intervals are gradually prolonged until conduction fails and the node recovers with a shorter conduction time This type is generally transient and treatment is seldom indicated A more serious type is the Mobitz type II in which random atrial impulses failto conduct This type is an indication that complet
24. e heart block might occur sometime later on It results from degenerative diseases in the conduction system Pacemaker is indicated 3rd Degree Block total block The conduction path has failed completely and the ventricles beat at a slow rate which is independent of the SA node and therefore also of any increased demand because of exercise The sufferer experiences weakness and may occasionally lose conciousness Stokes Adams attack Before the advent of pacemakers the prognosis was a 50 possibility of survival for one year from the first attack ECG characteristics This rhythm indicates a complete lack of conduction between the atria and the ventricles Atrial rhythm is regular ventricular rhythm is regular but slower than the atrial rate P waves bear no relationship to QRS complexes The QRS complex can be narrow or wide depending on the site that takes over the pacemaker function 3 1 2 Disorders of Impulse Formation Sick Sinus Syndrome A second group of conditions that lead to the prescription of pacemakers are dysfunctions of the sinus node The AV conduction is satisfactory but the atrial rate is too slow sinus bradycardia or widely variable While not fatal it degrades the patient s quality of life 3 1 3 Ventricular Tachycardia Ventricular tachycardias are life threatening in themselves and can lead into fatal ventricular fibrillation Treatment requires cardioversion However patients may not be within reach of one
25. ernal interference to a minimum Epameinondas Petropoulos 11 MSc in Electronics A Programmable Pacemaker University of Hertfordshire After a pulse has been generated the pacemaker enters an insensitive period refractory period This is to ensure that no inappropriate signal is interpreted as an R wave The refractory period for ventricle sensed pacemakers is 250 400 msec The definition NASPE BPEG gives for a VVI pacemaker is the following Inhibited ventricular pacing pacemaker stimuli occur at the escape interval in the absence of spontaneous sensed events In the presence of a spontaneous event provided it occurs after the pacemaker refractory period the pacing stimulus output is inhibited and the next pacing cycle interval begins to time out The behaviour of a VVI pacemaker is independent of atrial activity Ventricular contractions are synchronous with atrial ones to ensure an appropriate degree of filling The VVI pacemaker will often capture the ventricle before it has been able to fill adequately with blood As a result cardiac output will fall by as much as 15 This is called pacemaker syndrome It would be better for these cases to have a slower unpaced rhythm with satisfactory cardiac output 3 3 Modalities A large number of different pacemaker configurations are possible to suit individual problems Either the atrium or the ventricle or both may be paced and the same applies to any sensing for synchro
26. f Hertfordshire Chapter 2 Electrocardiography The purpose of this chapter is to provide a basic discussion of the physiological aspects behind pacing Most of the material presented here can be found in Aston For further information refer there or any other physiologic and or biomedical instrumentation textbook 2 1 Cardiac Anatomy and Function The function of the heart is that of a double pump working in unison The right heart receives unoxygenated blood from venous system and sends unoxygenated blood to lungs The left heart receives oxygenated blood from lungs and sends oxygenated blood to systemic circulation The heart is located between the lungs Its base consists of two atria and its apex consists of two ventricles see Figure 2 1 The ventricles are the pumping chambers Atrium Ventricles SA node Left bundle branch AV node Purkinje system Right Ui bundle branch Figure 2 1 The depolarisation path through heart Source Aston Epameinondas Petropoulos 5 MSc in Electronics A Programmable Pacemaker University of Hertfordshire 2 2 Biopotentials in the Heart The sinoatrial SA node beats at a rate of from 70 to 80 beats per minute bpm at rest the atrioventricular AV node and the bundle branch beat at lower rhythms The SA node normally determines the heart rate and causes stimulation of the other tissue before tt reaches its self pacing threshold The ability of the SA node to
27. g with our heart the organ that never rests Immersed into the body a pacemaker embodies the relationship of trust that has flourished between man and machine in the 20th century Of all the human machine relationships it is one out of few that humans are totally dependent upon The commercial aspects of that relationship are not new and_have been cleverly exploited However there is more to come with the privatisation of the healthcare services across Europe and the abolition of disturbing price caps_and restrictions from national governments It is believed that customers will start paying more money than they do now for extending their lifetime Based upon this and the fact that heart related deaths are only second to strokes it was decided to gain a deeper understanding ofthe aspects behind implantable devices A pacemaker is the ideal starting point as it has been established for 30 years with easy to grasp principles and objectives The most feasible aspect for implementation within the time limits and resources of an MSc project was to get involved with the software behind pacing 1 1 Aims and objectives The aim of the project is to design and built a software and hardware model of a programmable VVI pacemaker The overall objective to be achieved was to make the system fire a pulse whenever there is no QRS or when it occurs at very long intervals This has been done in practice for the last thirty years The new thing here is that it i
28. he escape interval Global time is incremented by one everytime the ISR is entered It resets whenever there is a new heartbeat The escape interval used is the most recent one updated The second condition is the wait variable evaluated against the escape interval The wait variable is incremented by one eachtime the ISR is entered Its difference with global time is that it resets when the pacing pulse duration expires If the two above conditions are true the output is set to the pacing pulse amplitude After that the pacing pulse duration is incremented The last if statement_is dealing with the termination of the pacing pulse If the pacing pulse duration has reached its predetermined limit 0 1 sec the pacing pulse stops Its duration variable resets as welas the wait variable The global time is not reset If on the next calls to the ISR there is not a heartbeat detected the wait signal ensures that the next pacing pulse will happen only when the most recent escape interval is exceeded again Pacing pulses do not last indefinitely because of the wait variable 7 3 3 Updating the Escape Interval The parameter that is taken in to account for deciding if it is appropriate to pace is the time difference between QRS complexes To measure time differences a clock is needed The internal programmable clocks could be used for this purpose but in order to minimise CPU load in setting up and running the clocks a simple
29. hm 30 180 Idioventricular 35 75 Ventricular Tachycardia 120 210 Ventricular Fibrillation Very coarse Very Fine Asystole 1 Atrio Ventricular Block 40 85 2 Atrio Ventricular Block Type 1 60 110 2 Atrio Ventricular Block Type2 60 110 3 Atrio Ventricular Block 30 50 Epameinondas Petropoulos 15 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Chapter 5 DSP Board and Development tools This chapter presents the features of the DSP starter kit which includes software and hardware It was used as the processing unit of the ECG signal The code name of the package is TMS320C6211 DSP Starter Kit produced by Texas Instruments and operating at 150 MHz Its components capabilities and limitations will be presented1n relation to the work done and as information for any further work 5 1 Introducing the TMS320C6211 C6211 The C6211 is a low cost version of the original C6000 device the C6201 The C6211 device provides 1200 MIPS million instructions persecond at 150 MHz Fixed point device It is based on very long instruction word VLITW architecture with flexibility to overcome problems of early VLIW designs That helps to have increased instruction level parallelism It has eight execution units These units operate in parallel They can perform up to eight instructions during a single clock cycle On chip peripherals include Exte
30. ibited G 80 Inhibited Activated As t can be seen the system has a problem recognising QRS types D and G They are both Bundle Branch Blocks Type D represents a Bundle Branch Block with broad R wave and inverted T wave see Figure 8 1 Type G represents a Bundle Branch Block with broad S wave and upright T wave see Figure 8 2 Epameinondas Petropoulos 33 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Figure 8 1 Normal Sinus Rhythm rate 80 QRS Type D Panah TEWUTA fi A Nin N FEY A T V V V V Figure 8 2 Normal Sinus Rhythm rate 80 QRS Type G Bundle branch blocks are disorders of the conduction system They cause delay of impulse propagation from the atria to the ventricles Bundle branch blocks involve failure of some portion of the intra ventricular conductionSystem They do not lengthen the P R interval as 1 block but the they delay the conduction of the pulse within the ventricles That is why the QRS is broad Depending on the particular site of infection either the Q R interval lengthens type D or the R S interval lengthens type G This is rarely a problem but leads to an abnormal ECG pattern that may make the detection quite difficult Two cases can cause the system not to detect a natural beat a The QRS is not of high enough amplitude to overcome the amplitude threshold and reset the pacing timer b The QRS is too broad and exceeds the refractory period
31. initiate an action potential without external stimulation is called automaticity The depolarisation of the SA node spreads through the atrium and reaches the AV node in about 40 msec Because of the low conduction velocity of the AV node tissue it requires about 110 msec for the depolarisation to reach the bundle branches The ventricles then contract the right ventricle forcing blood into the lungs the left ventricle pushing blood through the circulation system The contraction period of the heart is called systole The action potentials in the ventricle hold for 200 to 250 msec This relatively long time allows the ventricular contraction to force blood into the arteries The heart then depolarises during a rest period called diastole Then the cycle repeats The heart pumps around 4 8 litres minute 2 3 The Electrocardiogram During diastole while the heart 1s at rest all of the cells are polarised so that the potential inside each cell is negative with respect to the tissue not yet polarised Depolarisation occurs first at the SA node making the outside of the tissue negative with respect to the tissue not yet depolarised This imbalance results in an ionic current I causing the left arm LA to measure positive with respect to the right arm RA the resulting voltage is called the P wave After about 90 msec the atrium is completely depolarised and the ionic current measured by lead I reduces to zero The depolarisation then passes
32. into the right ventricular muscle depolarising it and making it negative relative to the still polarized left ventricular muscle Again the direction of current causes a positive voltage called the R wave The complete waveform is called an electrocardiogram ECG with labels P Q R S and T The P wave arises from depolarisation of the atrium The QRS complex arises from depolarisation of the ventricles The magnitude of the R wave within this complex is approximately 1 mV The T wave arises from repolarisation of the ventricle muscle Epameinondas Petropoulos 6 MSc in Electronics Sinus Node AV node Bundle Pacemaker 40 60 of His 70 80 bpm bpm Bundle Purkinje Ventricles Branches Fibres QRS amp T wave Figure 2 2 Normal Path of Impulse Conduction Table 2 1 lists the typical durations of ECG features These durations formed the basis of the timing cycles used in the software see Chapter 7 Table 2 1 Typical Duration of ECG features Feature Duration msec QRS complex 70 110 R R interval 600 to 1000 P R interval 150 to 200 S T interval 320 For reference purposes all ECG definitions are shown on figure 2 3 Epameinondas Petropoulos 7 MSc in Electronics A Programmable Pacemaker University of Hertfordshire R P R S T segment segment P R S T interval interval QRS complex QT i interval Intervals P R Beginning of P wave to beginning of QRS c
33. irown rhythm most of the time Worse a pacing pulse might fall into the T wave gt This could initiate tachycardia or even fibrillation A triggered circuit waits for a natural beat sensed electronically If it fails to appear after a set time an artificial pacing pulses output if a natural beat appears the device fires a pulse at the same time Triggered mode has the advantage that if the conduction path is only partly working it can boost it with its added pulse Also in the presence of interference it will continue to fire whereas an inhibited mode device could stop It must contain a refractory period after the pacing pulse to ensure interference does not cause an R on T pulse with its attendant danger of fibrillation Inhibited mode pacemakerts function only when sensing that the heart is not contracting They are inhibited by the presence of an R wave They count out a specified elapsed time interval following the R wave In the case of VVI pacemakers if no R wave appears before they time out a pulse is released to stimulate the ventricles If an R wave did appear during countdown the interval counter is reset The VVI pacemaker ensures a heartbeat no slower than its set rate Inhibited mode has the great merit that for many heart conditions it is only drawing current for the small fraction of the time that the heart s own rhythm is disturbed For this reason inhibited mode is preferred with care taken to keep the effects of ext
34. iting time for a depolarisation expires and invokes a pacing pulse Figure 8 18 shows the events and figure 8 19 shows the results Figure 8 18 NSR rate 80 PACs QRS Type A ee E S S Figure 8 amp 19Premature Atrial Complexes QRS Type A Rate 80 Epameinondas Petropoulos 47 MSc in Electronics A Programmable Pacemaker University of Hertfordshire 8 3 2 Premature Nodal Junctional Complexes PJCs Premature nodal complexes represent electrical impulses that occur at the AV node or high in the ventricles before the arrival of the next atrial beat The ECG reveals a P wave that may be prior to just after or during the QRS complex depending on the area of excitation Treatment is usually not required unless the rate is quite slow and coincident with a failed sinus mechanism The premature nodal waveforms available from Heartsim do involve anormal sinus rhythm hence treatment is not in theory required However because the QRS happens early in relation to the previous pulse it shortens the average interval The next QRS is expected from the pacemaker earlier than would otherwise be expected If the next QRS is not premature it will happen later than the calculated average interval and the pacemaker would fire a pulse Figure 8 20 shows a representative waveform and figure 8 21 shows the response of the pacemaker when applying such a waveform a7 eat EE oe ped a Figure 8 20 NSR rate 80 PJCs QRS Type A Figure 8 21 Prematu
35. limit Appendix 3 dss_cisr Calculate Interval section second argument on if statement GLOBAL _TIME gt 1000 Thus the same QRS is counted as two The interval in between them is very small falls out of the absolute limit Appendix 3 dss_cisr Calculate Interval section hysteresis gt 5000 and standard pacing rate takes place Appendix 3 dss_cisr Calculate Interval section else escape_interval 6000 The standard pacing rate is 60 bpm The second ease holds true for bundle branch blocks because they are too broad Tests were made with progressively increasing refractory period limit and types D and G were recognised That came with strings attached in that other waveforms were not recognised That holds for mild tachycardias which might be due to physical exercise In that case two pulses were interpreted as one because they were happening within the refractory blind period The output was half the rate Epameinondas Petropoulos 34 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Refractory period is the period that starts at the onset of the action potential during which a subsequent impulse cannot stimulate the cell 8 2 Basic Rhythms In this section all basic rhythms with all QRS types are tested to see if the pacemaker functions as expected under all circumstances 8 2 1 QRS Type A Type A QRS complex is a normal upright QRS T Figure 8 3 Ie i o a Figure
36. meinondas Petropoulos 31 MSc in Electronics A Programmable Pacemaker University of Hertfordshire 7 4 Shortcomings and Evaluation of Code Most of the discussion in this section concentrates on the source file dss_cisr which can be found on Appendix 3 There is were the intervention made to the example program to accommodate for the pacing algorithm It was done in the body of the ISR which serves the ADC ISRs in general should do something very quickly pass a value check a variable and then finish If they are too long they might cause problems with the next ISR orthey might not let other processes to run since they have the highest priority In the caseof this program processing within the ISR did not cause any conflict For two reasons The processing demands of the algorithm are low The DSP executes instructions very fast There are too many global variables in the code C theory states that global variables are bad practice due to reduced programmed readability because functions do not clearly indicate to the reader what parameters are passed into or out For two reasons global variables can be justified in the program Information needs to be retained in between calls to the ISR especially for timing variables Readability of functions is not reduced as only one function is accessing them and that function is on the same source file as the one that the global variables are declared Code was written in C so as to be transpotable t
37. nagement Device JTAG Controller Provides easy emulation and debugging Expansion Daughter Card Interface Provides extensible system development CE Compliant Universal Power Supply DSK vg gg sent ses sort lt siminishin oeisio Eis amw ie n Ex tinin Hiii wriraiyi wanay iai aa e s TD a raenswns ta erodi ta Ty 5 srs w fy at dpidtit sb s career teat ih omy URE A r vivant ere Figure 5 1 621 DSP board Epameinondas Petropoulos 17 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Software Included C6000 C Compiler and Assembly Optimizer Code Composer Debugger DSK support software Flash utility Sample programs including the audio example used in this project Confidence tests 5 3 Code Composer Studio Code Composer Studio CCS allows the developer toycreate build test and debug programs The program must be built loaded and then run Before loading the DSP must reset In the build options window there are Compiler Assembler and Linker tabs There are various commands for each of these fields Full optimisation on the compiler makes the program more efficient Not optimised code runs slower than optimised code When developing and testing programs itis often needed to check the value of a variable during execution Breakpoints and the Watch window can be used Use the step command to access instructions after the breakpoint Function
38. nd for sufficient rejection of interference a minimum CMRR of 100 dB is needed and an input impedance of the amplifier itself of at least 10 Q to prevent source impedance imbalances from deteriorating the overall performance As stated above the CMRR of the op amps used is 86 dB lower than what in theory is needed But because this project is not real practice though it has a real budget the 86 dB of the 0 30 TL071 were considered enough for the purpose of the project The results also justify the choice TLO71 input impedance is 10 Q amagnitude of a 1000 more than theory states as optimum In addition to electromagnetic interference noise generated by the amplifier and the connection between the source and the amplifier has to be taken into account High signal to noise ratios require the use of very low noise components Very low noise 15 nVVHz op amps were used to reduce the noise inherent in the amplifier The purpose of the high pass and low pass filters is to eliminate interference signals like amplifier offset potentials around 0 5V in our case and to reduce the noise amplitude by the limitation of the amplifier bandwidth Based on the ECG signal parameters amplitude of around 2 mV bandwidth of 1 100 Hz amplification factor of 200 and a bandwidth of 30 Hz 9 40 was chosen The highest frequency inthe ECG signal is contained in the R wave and is 10 30 Hz For diagnostic experimental or clinical applica
39. nisation In addition there are different responses to natural beats along with programming and communications options Table 3 1 shows a classifcation system as adopted by NASPE BPEG Epameinondas Petropoulos 12 MSc in Electronics A Programmable Pacemaker Table 3 1 NASPE BPEG Generic Pacemaker Code University of Hertfordshire Position I II MI IV V Category Chamber s Chamber s Response to Programmability Antitachy paced sensed sensing rate modulation afrythmia functions 0 none 0 none 0 none 0 none 0 none A Atrium A Atrium T P Simple Prog P Pacing Triggered anti tachy V V I M Multi prog S Shock Ventricle Ventricle Inhibited D Dual D Dual D Dual S D Dual A V A V T I Communication P S R Rate Modulation The choice of chambers to pace or sense depends on the individual pathology 2nd degree block might be dealt just by ventricular pacing while a heart suffering total block might need both chambers to be based and sensed Epameinondas Petropoulos 13 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Chapter 4 The Heart Simulator The Laerdal Heartsim 2000 is an ECG and Haemodynamic waveform simulator gt Only the ECG simulator is used It is designed for realistic simulations in teaching arrythmia recognition and defibrillation cardioversion The Heartsim 2000 consists of a centralunit with a
40. o as to observe its response together with the events This is not the case in real time pacemakers The system only outputs data when necessary The highest frequency in the ECG signal is about 30 Hz and is contained on the R wave According to the sampling theorem any continuous time signal must be sampled at least at twice its highest frequency so that it can be completely recovered from its samples The R wave is rapidly changing andvits amplitude is very important in determining whether a pulse has occurred It is best practice to be sampled at least five times its highest frequency at 150 Hz so the R wave amplitude is not undersampled and important information is lost 7 2 Utilising anexample program The following discussion describes the operation of the audio example which was included in the DSP package It can be found in detail in document SPRA598 An Audio Example Using DSP BIOS from the official Texas Instruments website www ti com The software that was written for the purpose of the project is presented after this discussion not to cause confusion of what was done and what was already written Epameinondas Petropoulos 25 MSc in Electronics A Programmable Pacemaker University of Hertfordshire The audio example explained next suits the application as it is desired to continuously sample and process data The audio example was used to move data from the ADC to the serial port and from there to a software in
41. o other devices Epameinondas Petropoulos 32 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Chapter 8 Results After discussing the algorithm in the previous chapter the presentation of the system is complete In this chapter the results obtained from the system are presented The approach followed is to test the system with all types and combinations of waveforms available from Heartsim Results were observed on TCSCOPE Natural beats and pacing pulses appear on the same graph Graphs are shown selectively to clarify responses in difficult to explain combinations of waveforms The general rule is that the pacemaker should respond to natural waveforms of rate below 60 bpm by firing a pacing pulse Pacing pulse width is deliberately wider to help visualization on screen and distinguish it from natural beats In the results presentation emphasis is given on sequences that require pacemaker treatment 8 1 QRS Types The system was tested with the different QRS types provided by the Heartsim running at their default rate of 80 beats per minute bpm to check if it performs what was expected In all cases it was expected that the pacemaker be inhibited at 80 bpm Table 8 1 QRS Types and Pacemaker Response QRS Type Rate default Expected Result A 80 Inhibited Inhibited B 80 Inhibited Inhibited C 80 Inhibited Inhibited D 80 Inhibited Activated E 80 Inhibited Inhibited F 80 Inhibited Inh
42. omplex S T End of S wave to end of T wave Q T Beginning of Q wave to end of T wave Segments P R End of P wave to beginning of Q wave S T End of S wave to beginning of T wave Complex QRS Beginning of Q wave to end of S wave Durations Average durations shown on drawing in seconds Figure 2 3 ECG definitions Source Ast n Epameinondas Petropoulos 8 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Chapter 3 Pacemakers In this chapter the physiologic conditions under which pacing is indicated are presented together with the different types of pacemakers Most of the discussion is taken from lecture notes prepared by Mr Tony Compton for the Medical Electronics course and by the Webster encyclopaedia 3 1 Indications for Pacing In chapter 2 it was shown that the rhythm of a healthy hearts determined by spontaneous pulses generated by the sino atrial SA node Signals from the SA node spread through the atria causing them to contract are delayed in the atrio ventricular AV node conveyed rapidly through the His Purkinje fibres and finally activate the ventricular muscle 3 1 1 Disorders of Impulse Conduction Conduction Delay and AV Block If any part of the impulse conduction path 1s faulty see Figure 2 2 the condition of heart block occurs This occurs in several degrees of severity The most usual dysfunction of this path happens on the bundle of His 1st Degree Block The
43. r approach was taken The ISR was Epameinondas Petropoulos 29 MSc in Electronics A Programmable Pacemaker University of Hertfordshire used as an indicative clock It occurs at regular intervals and has the highest priority With the utilisation of global variables numerical values were saved in between the calls to the ISR Time is measured as number of calls to the ISR 500 calls equal 0 1 sec 5000 calls equal sec When the program recognises a heartbeat that does not happen within the refractory period it is given the green light to update the new average escape interval The last three intervals are averaged One of them each time is updated A hysteresis period is added to the detected average value to accommodate for normal heart rate variability This is set to plus ten percent The detected escape interval is then compared to two limits If the average interval exceeds any of those limits the interval is set at a predetermined value of 1 sec If it is within the limits its value is the one that was calculated The long interval limit 1 2 sec purpose is to set a maximum waiting interval Thus if the detected average beat interval exceeds 1 2 sec then the program should not assign the calculated value to the escape interval It is evident that the heart is experiencing a serious problem The predetermined interval is set 1 sec It is short enough to illicit a pacing pulse Hence a hard limit is established Heartbeats occurring in longer
44. ral solution to the problem involves voltage followers to attain high input impedances The two input op amps 3 and 4 ir figure 5 1 provide high differential gain and unity common mode gain without the requirement of close resistor matching The differential output from the first stage represents a signal with substantial relative reduction of the common mode signal and is used to drive a standard differential amplifier which further reduces the common mode signal The gain of the instrumentation amplifier is shown in equation 6 3 3 3 pae 2x R2 ae R5 2 x 68 x 10 ely zW 200 6 3 RI RA 680 10x10 The pre amplifier provides additional gain due to the very low level of the ECG signal The pre amplifier stage is badly placed before the instrumentation amplifier It cancels the advantage of high input impedance Any further amplification to the signal should be provided after the instrumentation amplifier In the context of the project it did not make muchdifference to the signal It was placed there in order to avoid having to built the amplifier from scratch The instrumentation amplifier and the filter had been built before realising that further amplification was needed It is bad practice and should be avoided Epameinondas Petropoulos 23 MSc in Electronics A Programmable Pacemaker University of Hertfordshire TLO 1 nol LA ECG Amplifier Figure 5 1 ECG Amplifier Schematic diagram produced using
45. re Junctional Complexes QRS Type A Rate 80 Natural beats appear negative Pacing pulses appear positive 8 3 3 Premature Ventricular Complexes PVCs Premature ventricular beats represent ventricular depolarisation that prematurely arises from a site in the ventricular muscle or the conductive system If the site of impulse initiation is within the common section of specialised conductive tissue the QRS complex willbe slightly widened since it reflects retrograde and anterograde conduction within the Epameinondas Petropoulos 48 MSc in Electronics A Programmable Pacemaker University of Hertfordshire ventricles simultaneously The ECG reveals wide bizarre QRS complexes PVCs can signal an increased risk of a more dangerous type of ventricular conduction disturbance Figure 8 22 shows a representative waveform and figure 8 23 shows the response of the pacemaker wh n applying such a waveform Ly poly pot Figure 8 22 NSR rate 80 Unifocal PVCs QRS Type A Figure 8 23 Unifocal Premature Ventricular Complexes QRS Type A Rate 80 Pacing pulses appear positive and wider Section 8 4 Artefacts and Noise The Heartsim offers the opportunity to test the pacemaker under conditions of noise superimposed on the ECG see Figure 8 24 Two kinds of noise are offered 60 Hz AC line interference from other machinery and muscular interference from patient movement These were superimposed ona variety of waveforms both on their o
46. ricular Fibrillation Rapid irregular rhythm with chaotic undulations without a P wave QRS ST segment or T wave see Figure 8 5 Itrepresents lack of organised ventricular depolarisation Result Pacemaker is active It does not interpret any part of the waveform as a QRS complex correctly It fires a pacing pulse at regular intervals It is not a substitute to a cardioverter defibrillator but it can in some cases terminate the fibrillation wry m ann ws Figure_8 5 Coarse Ventricular Fibrillation Epameinondas Petropoulos 37 MSc in Electronics A Programmable Pacemaker 8 2 2 QRS Type B University of Hertfordshire Type B QRS complex is an upright QRS with ST depression and T inversion Table 8 3 QRS Type B in different rhythms and rates Rhythm Rate Expected Result Sinus Bradycardia 30 Activated Activated 40 Activated Activated 50 Activated Activated Normal Sinus 60 Inhibited Inhibited Rhythm 80 Inhibited Inhibited 100 Inhibited Inhibited S Tachycardia 120 Inhibited Inhibited A Tachycardia 180 Inhibited Inhibited A Fibrillation 90 Inhibited Inhibited Junctional 80 Inhibited Inhibited V Tachycardia 180 Inhibited Activated V Fibrillation Activated Activated lo AV block 40 Activated Activated 60 Inhibited Inhibited 85 Inhibited Inhibited 20 AV block 1 5 4 note 1 1 paced 4 naturals 1 paced 4 naturals 4 3 1 paced 3 naturals 1 pa
47. rnal Memory Interface Can interface to 8 16 and 32 bit SDRAM and asynchronous memories Multi Channel Buffered Serial Ports McBSPs One of them is used in the project Serial Peripheral Interface SPI compatible Host Peripheral Interface 16 bitasynchronous Two 32 bit Timers None of them is used Interrupt Selection McBSP1 Receive interrupt is used There are other 15 hardware interrupts to use The C6211 is a cache based device On chip cache provides a 98 hit rate The cache architecture allows for the low cost of the device Saves up by using slow less expensive memory On chip memory consists of 72 Kbytes 8 Kbytes serving as level 1 cache that CPU can directly access Of that 4 Kbytes are used for program and 4 Kbytes for data The data path to the CPU allows for eight 32 bit instructions during every cycle The cache uses a Epameinondas Petropoulos 16 MSc in Electronics A Programmable Pacemaker University of Hertfordshire least recently used replacement scheme The remaining 64 Kbytes of on chip memory canbe used as level 2 cache or be directly mapped as internal memory RAM 5 2 C6211 Contents Hardware included C6211 DSK Board Easily connects to a PC through a parallel port cable included 150 MHZ C6211 DSP see Figure 6 1 4 MB External SDRAM and 128 KB External Flash Provides additional program and data storage TI S TLC320AD535 16 bit Data Converter TI S TPS56100 Power Ma
48. s implemented on a Digital Signal Processor DSP giving the advantage of even more processing powervat higher speeds if needed an advantage that was not fully utilised by this work mostly due to time constraints and inexperience However established ideas were implemented in a new and fast machine like the Texas Instruments C6211 DSP setting a useful basis for further work to be done by another student Epameinondas Petropoulos 1 MSc in Electronics A Programmable Pacemaker University of Hertfordshire 1 2 Overview of the project The signal that is supposed to be the heart signal was taken from a heart simulator Laerdal Heartsim2000 This has an extensive electrocardiogram ECG library withvery accurate representations of seven QRS types thirteen basic rhythms and five ventricular premature complexes The signal being a realistic ECG output was of very small amplitude of around 2 mV pk pk and needed amplification An amplifier was therefore build for three purposes as an anti alias measure for the interface with the analog to digital converter ADC an amplifier and filter for the ECG signal and also to couple with the very high impedance of the input signal The amplifier consisted of two op amps acting as pre amplifiers an instrumentation amplifier and a Bessel 2nd order bandpass filter 9 40 Hz The amplifier output was connected to the DSP via an ADC and serial port At the processing stage the ECG signal was stripped out
49. s can be stepped over or can be stepped into for detail Structures can be monitored on the watch window by expanding the structure name in the window CCS also offers code profiling features to gather statistics about the execution of any part of the program The profiling clock counts instruction cycles Profile points report the number of instruction cycles since the previous profile point or since the program started running On the profiling statistics window the feature that is most important is the minimum number of cycles Graphs can be added to view the signal in time or frequency domain This option comes handy when using digital filters Animation of code is also on offer if needed During animation a coloured line over the statement highlighted switches on and off as the program executes that statement All the above mentioned features while helpful in debugging the code they consume many cycles to execute When it is decided that the program is bug free these features should be switched off Otherwise data might belost especially when using I O CCS also offers the DSP BIOS tools These include the statistics windows used with the profiling points The execution graph shows in real time the execution of all processes and threads in the program How many clock cycles each process consumes and if there is Epameinondas Petropoulos 18 MSc in Electronics A Programmable Pacemaker University of Hertfordshire any conflict between the processes
50. ted Hence the operation of the pacemaker under such a waveform is considered successful Figure 8 9 3 1 conduction ratio of 2 AV block Type 2 Natural beats appear inverted Pacing pulses appear positive and wider Epameinondas Petropoulos 40 MSc in Electronics A Programmable Pacemaker 8 2 3 QRS Type C University of Hertfordshire Type C QRS complex is an upright QRS with ST elevation see Figure 8 10 Secchi L TML wa N Figure 8 10 Normal Sinus Rhythm rate 80 QRS Type C Table 8 4 QRS Type C in different rhythms and rates Er fee TDR eS TE Rhythm Rate Expected Result Sinus Bradycardia 30 Activated Activated 40 Activated Activated 50 Activated Activated Normal Sinus 60 Inhibited Inhibited Rhythm 80 Inhibited Inhibited 100 Inhibited Inhibited S Tachycardia 120 Inhibited Inhibited A Tachycardia 180 Inhibited Inhibited A Fibrillation 90 Inhibited Inhibited Junctional 80 Inhibited Inhibited V Tachycardia 180 Activated Activated V Fibrillation Activated Activated lo AV block 40 Activated Activated 60 Inhibited Inhibited 85 Inhibited Inhibited 20 AV block 1 5 4 1 paced 4 naturals 1 paced 4 naturals 4 3 1 paced 3 naturals 1 paced 3 naturals 3 2 1 paced 2 naturals 1 paced 2 naturals 20 AV block 2 3 1 2 paced 1 natural 2 paced 1 natural 3 1 2 1 interchanged interchanged 2 1 1 paced 1 natural 1 paced 1
51. terrupt function and back againto the serial port connected to the ADC The aim of the example is to use the DSP BIOS TI kernel for scheduling data transfer between the hardware I O peripherals ADC and the DSP In short for managing data transfer The DSP BIOS pipes or PIP module was used to buffer streams of program input and output data They manage I O in blocks also called stream based or asynchronous I O Each pipe buffer is divided into a number of frames The pipe has two ends the writer and the reader Data notification functions are performed to synchronisedata transfer One end of the pipe is controlled by the hardware ISR serial port receiver ISR and the other end is controlled by the software interrupt function Data input from the codec flows from the serial port ISR through the receiver pipe DSS_rxPipe to the software interrupt where it4s copied_to the transmitter output DSS _txPipe and sent back to the serial port to be transmitted through the codec It is on the bit of the ISR that writes the sample value to the Data Transmit Register that intervention was made to modify the data so as to monitor and control the output at our will Epameinondas Petropoulos 26 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Software Interrupt Transmit Frame Receive Frame Serial Port Receive amp Transmit Interrupts Service Routines DRR Serial Port DXR Analog to Digital Converter
52. than the hard limit intervals will activate pacing pulses Also when there is no ventricular contraction at all the pacemaker will be active in its default rate 1 sec Detected average intervals that breach the short interval limit 0 8 sec indicate that the heart rate is high enough not to need constant supervision Thus the average waiting interval is set at the predetermined rate which is much longer If a short limit was not in place the program would follow a heart rate of say 100 bpm and when it slowed down say below 90 bpm it would start firing a pacing pulse which is undesirable In detected average intervals that are within the limits the escape interval essentially traces the heart rate The program is like being in an alert state Epameinondas Petropoulos 30 MSc in Electronics A Programmable Pacemaker University of Hertfordshire update Global Time LOW Threshold Detection HIGH NO YES Is event sensed AND Global time larger than the refractorv nerind Use previous Interval Update Average Interval add Hysteresis and reset Global time Evaluate Average Interval Information within limits out of limits Escape Interval TN Default Escape Average Interval wp Interval plus Hysteresis SIN SL Is Global time bigger than Escape Interval YES Ci eee NO PACE Get Next Sample Is Pacing Pulse Duration finished NO EB or Figure 7 3 The pacemaker algorithm Epa
53. the pacemaker would fire one pulse for every natural that is occurring In figure 8 17 the results are displayed The pacing pulses fire regular as the natural beats Theypacemaker replaces the dropped beat through the sequence Operation is considered successful in this case Ideally the pacing pulses should happen in the middle of the intervalof the two natural beats Epameinondas Petropoulos 46 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Figure 8 17 2nd degree block type 2 2 1 Natural beats appear positive and narrow Pacing pulses appear positive and wider Vertical axis units are volts Horizontal axis units are seconds 8 3 Premature Beats 8 3 1 Premature Atrial Complexes PACs Premature atrial complexes originate within the atria muscle but outside of the SA node from an ectopic focus The ECGs P wave is premature but the QRS complex remains unchanged because AV nodal conduction and ventricular depolarisation Unless the heart rate is so rapid that the ventricles do not have time to fill properly PACs are not considered a dangerous disarrythmia The VVI pacemaker senses the ventricular depolarisation calculates the average interval of the three previous pulses and reduces it if the beat occurs prematurely The next normal ventricular depolarisation happens atan intervalthat is longer than the average interval of the three previous beats which includes the short premature interval The wa
54. tion part of the algorithm and why under laboratory conditions it proved so reliable is discussed The critical part of the pacemaker timing cycles is discussed in full detail When to start and stop pacing together with updating intervals of pulses and calculating their average interval are matters on which the discussion develops A flowchart accompanies the discussion In Chapter 7 presentation of the work done during the project ends Chapter 8 presents the results obtained from the system QRS types on their default rhythms are tested first and there is evidence that the system does not recognise bundle branch blocks All other types are recognised successfully Then for each QRS type successfully recognised all the basic rhythms available are presented as inputs to the system The system is found to respond in line with the expectations in all cases When pacing is needed it is activated and inhibited when it is not Emphasis is given on AV blocks Their conduction ratios and the responses of the pacemaker are explained in detail Graphs are included when necessary to help the reader visualise the processes and timings involved Premature beats are then tested as inputs and they illicit a single pacing pulse Their physiologic origin is briefly explained And finally waveforms with noise are presented and the system output is found undistorted and satisfactory Epameinondas Petropoulos 4 MSc in Electronics A Programmable Pacemaker University o
55. tions the bandwidths employed differ and are somewhat larger than the 30 Hz employed here The T wave contains no necessary information for the pacemaker Its frequency is approximately 5 Hz or less The lower cut off frequency of the amplifier is 9 Hz and most of the T wave energy is cut off Epameinondas Petropoulos 22 MSc in Electronics A Programmable Pacemaker University of Hertfordshire 6 3 The Amplifier The ECG amplifier design is shown in Figure 5 1 It consists of three stages The pre amplifier op amps and 2 the instrumentation amplifier op amps 3 5 and the bandpass filter op amps 6 and 7 The bandpass filter is needed to cut off frequencies outside the useful ECG bandwidth and to reduce the effect of aliasing induced by sampling Without filtering the 50 Hz mains noise or other noise could replicate itself in the signal band corrupting the information It is a second order Bessel roll off Its cut off frequencies are shown on equations 6 1 and 6 2 1 1 InRC 2r x 220x10 x82 x10 1 1 RC 2m x 150x107 x 27K10 Lower cut off frequency fic 8 82 Hz 6 1 Upper cut off frequency fac 39 2Hz 6 2 The main task of the instrumentation amplifier is to sense the voltage between the measuring electrodes RA LA while rejecting the common mode signal Such a differential amplifier cannot be realized using a single op amp design since this does not provide the necessary high input impedance The gene
56. tivated Very Fine Activated Activated Asystole Activated Activated 1 AV block 40 Activated Activated 60 Inhibited Inhibited 85 Inhibited Inhibited 2 AV block 1 5 4 slow 1 paced 4 natural 1 paced 4 natural 4 3 note 3 note 3 3 2 fast 1 paced 2 natural 1 paced 2 natural 20 AV block 2 3 1 slow note 4 note 4 Y3N interchanged interchanged 2 1 fast 1 paced 1 natural 1 paced 1 natural 30 AV block 30 Activated Activated 50 Activated Activated 60 Inhibited Inhibited Epameinondas Petropoulos 36 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Notes 1 Ventricular Tachycardia Defined as three or more consecutive ventricular ectopic beats see Figure 8 4 The QRS is broad and different from supraventricular complexes Rhythm is regular or nearly regular Atrial rhythm is usually slower and undisturbed by the rhythm of the ventricles Sustained ventricular tachycardia is a life threatening arrhythmia and definitive treatment should be instituted immediately The programming of the pacemaker does not include any anti tachy arrhythmia functions However it fires pacing pulses when it has as input ventricular tachycardias waveforms That is desirable as a treatment Pacemaker pulses properly timed and placed can terminate ventricular tachycardia Such treatment can also elicit fibrillation AAAA AAAA AAAA Y Figure 8 4 Ventricular Tachycardia rate 180 QRS type A 2 Vent
57. to 1 conducted ORS These ratios are used interchangeably Figure 8 13 shows a representative waveform of QRS type E a j i aa L ie a A Figure 8 13 2nd degree block type 2 3 1 2 1 conduction ratio Source Heartsim 2000 User Manual It is expected to get 2 paced pulses for 1 natural beat and 1 paced pulse for 1 natural beat Some normal pulses appear in between the block sequences as would happen in a real heart it is not always in a state of block Figure 8 4 shows the result The pacemaker pulses successfully replace all the dropped beats at reasonable intervals from the last natural beat Figure 8 14 2nd degree block type 2 3 1 2 1 conduction ratio Natural beats appear biphasic last one appears positive All paced pulses appear positive and wider Vertical axis units are volts Horizontal axis units are seconds Epameinondas Petropoulos 44 MSc in Electronics A Programmable Pacemaker 8 2 5 QRS Type F University of Hertfordshire Type F QRS complex is a QRS with ST elevation Table 8 6 QRS Type F in different rhythms and rates Rhythm Rate Expected Result Sinus Bradycardia 30 Activated Activated 40 Activated Activated 50 Activated Activated Normal Sinus 60 Inhibited Inhibited Rhythm 80 Inhibited Inhibited 100 Inhibited Inhibited S Tachycardia 120 Inhibited Inhibited A Tachycardia 180 Inhibited Inhibited A
58. ue to physiological and or instrumentation reasons Because of that both the amplitude and its duration are needed to testify a pulse However the output from Heartsim has a constant QRS amplitude Thus with certainty minimum amplitude can be predetermined and set at the right level Epameinondas Petropoulos 28 MSc in Electronics A Programmable Pacemaker University of Hertfordshire The system needs to be safeguarded against pulses happening at the refractory period and recognising them as ventricular depolarisation Such pulses are not taken into account when calculating the escape interval For the escape interval to be updated it needs both a sample above threshold and some time 0 2 sec to have passed since the last onset detection Samples above minimum threshold happening within the 0 2 sec time after the onset will appear at the output but they will not trigger an interval update Such samples might be the rest of the QRS pulse and maybe some false detected action potentials The stimulus duration and the refractory period are folded into one block of time 7 3 2 Pacing Decision If the input sample is below threshold not a naturalpulse and there is no pacing taking place at that moment the output is set to zero After that there is no updating of the escape interval The program reaches the point where itis to decide if a pacing pulse should fire or not It evaluates two conditions The first tests if the global time has exceeded t
59. ust be found since Heartsim gives only idealistic waveforms The standard ECG libraries that have been developed form the standard test for all implantable devices One thing that might need changing in the algorithm is to add variable threshold instead of the constant employed here Maybe a more sophisticated approach for threshold detection is needed such as the time averaging or successive high pass filters 9 5 Conclusion Pacemaker technology is a well established field for a couple of decades now This project did not intend to be a cornerstone on human knowledge It was rather a means to an end a basis for further work Arrhythmia recognition algorithms are an evolving field Before the project started it was hoped that there would be some time left to touch on the subject Probability Density Function for a start That is why a fast DSP was chosen But the Epameinondas Petropoulos 52 MSc in Electronics A Programmable Pacemaker University of Hertfordshire specific DSP was responsible for considerable delay Time was lost troubleshooting things not directly related to the project The pacemaker presented in this report might not be the presented the most exciting thing you have seen but at least it worked Epameinondas Petropoulos 53 MSc in Electronics A Programmable Pacemaker University of Hertfordshire REFERENCES 1 Aston Principles of Biomedical Instrumentation Chapter 2 2 Compton AJ Therapeutic Devices Sept 1991
60. vates a pacing pulse Because of the hysteresis some flexibility is allowed so as to accommodate for normal heart variability The output of the system is displayed together with the QRS Results show rigid performance The pacemaker is activated whenever heart rate is below 60 bpm and or whenever a single natural pulse does not appear in less than a second from the previous one The pacemaker is activated in very fast ventricular fibrillation and tachycardia although not programmed to The system does not recognise bundle branch blocks The system proved noise tolerant Premature complexes illicit a single pacing pulse In 2 AV blocks the pacemaker is activated when beats are missed and inhibited when beats were present In 3 AV blocks the pacemaker is always active In all rhythms that the pacemaker was expected to remain inhibited did so 9 2 Objectives and how they were met The objective that was set to achieve was for the system to fire a pacing pulse whenever there is no QRS or when it occurs at very long intervals The objective was met in a very high degree Pacing is activated whenever the heart rate is below 60 bpm and or whenever a single natural pulse fails to appear within about 1 second For VVI modalities the response of the pacemaker on 20 AV block is critical as it is under that condition that VVI pacemakers are prescribed The response of the pacemaker in 20 AV blocks Type 1 and 2 was excellent as it can be seen from sections 8
61. wer by not executing the ISR every time jt is called The least sampling rate that the C6211 can get down to is 22 kHz Because the project did not deal with real implantable devices it did not employ any of the above measures but the difference must be mentioned One way not to execute the ISR every time it is called is shown in Figure 7 2 int GLOBAL VARIABLE 0 void ISR void GLOBAL VARIABLE 1 if GLOBAL VARIABLE DELAY execute isr statements GLOBAL VARIABLE 0 j Figure 7 2 Virtual Slower Sampling Rate 7 3 The Pacing Algorithm 7 3 1 The Threshold Stimulation Only the QRS complex is of interest for a VVI pacemaker so the ECG signal was stripped out of everything by using a threshold level The threshold stimulation is defined as the minimum amplitude and stimulus duration that consistently results in cardiac depolarisation outside the physiologic refractory period From the definition it is evident that three parameters in a pacemaker design should be taken into account when detecting the minimum threshold of stimulation The minimum amplitude the stimulus duration and the refractory period The minimum amplitude is used to define if an input sample is part of a heartbeat or not This is a sample by sample approach No consideration about pulse duration is made at this point The pulseis recognised at its onset This is an idealistic approach In real world data capturing apparatus pulse amplitude may vary d
62. wn and combined All the QRS s were acquired with no distortion at all The system proved noise resistant The more obvious explanation is the high level of amplitude threshold Noise was much lower than that threshold However that might not be the case in a real world pacemaker The bandpass filter of the amplifier eliminates any baseline drift effects Epameinondas Petropoulos 49 MSc in Electronics A Programmable Pacemaker University of Hertfordshire ARTIFACTS NSR rte 80 with He AC line interference other machinery halaan NSR rate 80 with muscular imuesiesenoe from movement of the Figure 8 24 60 Hz AC line interference and muscular noise from Heartsim Source Heartsim 2000 User Manual Epameinondas Petropoulos 50 MSc in Electronics A Programmable Pacemaker University of Hertfordshire Chapter 9 Conclusions and Discussion 9 1 Summary of work and results The signal was acquired from the Heartsim 2000 ECG simulator It was amplified with an instrumentation amplifier and filtered using a 2nd order Bessel bandpass filter 9 40 Hz bandwidth Through a 16 bit ADC and a serial port it was fed into the DSP for processing The first step of processing is a threshold detector so that only the QRS information remains from the ECG signal The average time interval between heartbeats is calculated Some hysteresis is added to that interval If a pulse does not appear within one second the algorithm acti

Acknoweldgments - Introduction

Contents

Download Pdf Manuals

Related Search

Related Contents