Early Myocardial Infarction Detection
By Kasturi Joshi Edward Labrador (Team # 17)
A Project Report Presented to The Faculty of Department of General Engineering San Jose State University
In Partial Fulfillment Of the Requirements for the Degree Master of Science in Engineering May 2009
Early Myocardial Infarction Detection ii
© 2009 Kasturi Joshi Edward Labrador
ALL RIGHTS RESERVED
Early Myocardial Infarction Detection iii
APPROVED FOR THE DEPARTMENT OF GENERAL ENGINEERING
Dr. Leonard Wesley
Dr. Mallika Keralapura
Dr. Sudhi Gautam
APPROVED FOR THE UNIVERSITY
Early Myocardial Infarction Detection iv Abstract Cardiovascular heart disease, such as myocardial infarction, is the number one leading cause of death in United States. Having the ability to detect the symptoms and the ability to detect the onset of myocardial infarction can greatly decrease the mortality and morbidity of patients. This project report presents the ability of detecting the onset of symptoms of myocardial infarction using electrocardiogram (ECG). The proposed technique for identifying the isoelectric STsegment of an ECG is by using Biorthogonal Wavelet Transform. The ST-segment is then compared to an isoelectric baseline, using the PQ segment, which determines if there’s a presence of myocardial infarction. An ST-segment that deviates from the baseline by ±1mV is a probable myocardial infarction. Having an ST elevation for more than 5 minutes determines that a myocardial infarction is present and a patient needs to be alerted. A program based on Matlab software was written to perform the identification of myocardial infarction. ECG datasets were gathered from Physiobank’s Automated Teller Machine database. The accuracy of the written code and its ability to detect true positive myocardial infarction was determined using the ROC analysis. The performance of the code showed that it can accurately determine true positives and true negatives in an ECG dataset. The accuracy of this project was proven to approximately 73% from the 54 ECG datasets tested from 5 different physiobank databases.
Early Myocardial Infarction Detection v ACKNOWLEDGEMENT
We would like to express our gratitude to Prof. Dr. Mallika Keralapura, Dept of Electrical Engineering, San Jose State University and Dr. Sudhi Gautam for their generous guidance, encouragement, direction and support in completing this project.
We would like extend our sincere gratitude to Prof. Leonard P. Wesley, Dept. of Computer Engineering for an opportunity to pursue ENGR 298 course under his guidance, precious suggestions, and advice.
We would also like to extend our special thanks to the members of our family. Without their support and encouragement this project would not be complete.
- Kasturi Joshi - Edward Labrador
Early Myocardial Infarction Detection vi Table of Contents List of Figures .............................................................................................................................. viii List of Tables.................................................................................................................................. ix List of Equations ............................................................................................................................. x I.Objective ....................................................................................................................................... 1 II.Introduction ................................................................................................................................. 1 III.Anatomy of Heart ....................................................................................................................... 2 IV.Physiology of Heart ................................................................................................................... 5 V.Myocardial Infarction .................................................................................................................. 7 a.Diagnostic Studies to detect Myocardial Infarction............................................................... 11 i.Blood Analysis ..................................................................................................................... 11 ii.Imaging ............................................................................................................................... 14 iii.Electrical Activity Monitoring ........................................................................................... 16 VI.Electrocardiography ................................................................................................................. 16 a.Measuring ECG ...................................................................................................................... 18 VII.Wavelet Transforms ............................................................................................................... 23 VIII.Introduction to Early Myocardial Infarction Detection System ............................................ 33 a.Background............................................................................................................................. 33 b.Materials and Method ............................................................................................................ 36 i.Database Description .......................................................................................................... 36 ii.Method for ST-elevation detection ..................................................................................... 37 IX.Testing and Verification .......................................................................................................... 47 a.Results..................................................................................................................................... 47 b.Discussion............................................................................................................................... 48 X.Economic Justification .............................................................................................................. 50 a.Executive Summary ................................................................................................................ 50 b.Problem Statement.................................................................................................................. 52 c.Solution and Value Proposition .............................................................................................. 52 d.Market Size ............................................................................................................................. 53 e.Competitors............................................................................................................................. 55 f.Customers ................................................................................................................................ 56
Early Myocardial Infarction Detection vii g.Cost ......................................................................................................................................... 57 i.Fixed costs ........................................................................................................................... 57 ii.Variable Costs .................................................................................................................... 58 h.Price Point .............................................................................................................................. 59 i.SWOT Assessment ................................................................................................................... 59 j.Investment Capital Requirement ............................................................................................. 60 k.Personnel ................................................................................................................................ 62 l.Business and Revenue Model .................................................................................................. 63 m.Strategic Alliances/Partners .................................................................................................. 64 n.Profit and Loss ....................................................................................................................... 64 i.Demand Assumptions........................................................................................................... 65 ii.Product Assumptions .......................................................................................................... 65 o.Exit Strategy ........................................................................................................................... 66 XI.Future Directions ..................................................................................................................... 67 XII.Conclusion .............................................................................................................................. 68 XIII.References ............................................................................................................................. 70 Appendix A ................................................................................................................................... 74 Appendix B ................................................................................................................................... 87 Appendix C ................................................................................................................................. 110 Appendix D ................................................................................................................................. 112
Early Myocardial Infarction Detection viii List of Figures Figure 1: Anatomy of Heart [Source: Heart Information Center (2006)] ...................................... 3 Figure 2: Electrical Conduction System of the Heart ‘Bundle Branch Block’ of Heart [Source: Heart Information Center (2006)] ................................................................................................... 6 Figure 3: Myocardial Infarction [Source: Coronary Artery Disease, 2008]................................... 8 Figure 4: Resulting zones from Myocardial Infarction [Source: Myocardial Infarction (2009)] .. 9 Figure 5: An ECG with the major peaks and intervals. (Vibes Electrocardiogram, n.d.) ........... 19 Figure 6: Illustrates the cause of deflection of an ECG (O’ Grady, M.R., n.d.).......................... 19 Figure 7: An Einthoven's triangle with Lead I, II, and III. .......................................................... 20 Figure 8: Axial representation of Lead I, II, III, aVR, aVL, and aVF. (O'Grady, M.R., n.d.) .... 21 Figure 9: Typical ECG waveform [Source: Jouck. P.P.H. (2004)] .............................................. 22 Figure 10: 2.4 Biorthogonal Wavelet and ECG Signal ............................................................... 31 Figure 11: Types of Biorthogonal Wavelets available in Wavelet Toolbox 3.0 of Matlab 7.1 ... 32 Figure 12: Normal ECG waveform on Strip Chart [Source: Barron Jon, 2007] .......................... 34 Figure 13: Dyadic Wavelet Transform of ECG signal [Source: Jouck. P.P.H. (2004)] ............... 38 Figure 14: Biorthogonal Wavelet Transform of ECG Signal from 21 to24 level ......................... 39 Figure 15: Method for ECG Parameters Detection [Source: Tompkins, 2000] ........................... 40 Figure 16: Filter expressed in Direct Form II transposed structure [Source: Matlab7.1R14 Help] ....................................................................................................................................................... 41 Figure 17: Baseline Wander elimination ...................................................................................... 42 Figure 18: R-peak Detection and PQSTJK extraction of ECG wave at level 24 .......................... 43 Figure 19: Shows an intuitive GUI result for an ECG data with no MI detected......................... 45 Figure 20: Shows an intuitive GUI result for an ECG data with an MI detected. ........................ 45 Figure 21: Flowchart for ST-elevation detection ......................................................................... 46 Figure 22: Receiver Operating Characteristic Curve ................................................................... 49 Figure 23: An estimate of myocardial infarction prevalence in the United States ....................... 53 Figure 24: An estimate of new and recurrent incidence of myocardial infarction in the United States ............................................................................................................................................. 54 Figure 25: The direct and indirect cost of myocardial infarction per year ................................... 54 Figure 26: Initial Investment Requirement of MI Detector device .............................................. 61 Figure 27: Yearly Model of MI Detector device .......................................................................... 61 Figure 28: The quarterly model of SMART Medical Devices ..................................................... 66
Early Myocardial Infarction Detection ix List of Tables Table 1: Macroscopic & Microscopic Findings of MI [Source: Klatt.E.C, 2008] ....................... 12 Table 2: Typical Amplitudes and Durations for ECG signal [Source: Saritha. et.al. (2008)] ...... 35 Table 3: Testing Results for ST-change detection program ......................................................... 48 Table 4: Data Points for ROC Curve ............................................................................................ 49 Table 5: Fixed cost of MI Detector device ................................................................................... 58 Table 6: Variable cost of MI detector device ................................................................................ 59 Table 7: SWOT assessment of MI Detector medical device ........................................................ 60 Table 8: The break-even table of MI Detector device .................................................................. 62 Table 9: The quarterly model of SMART Medical Devices ......................................................... 66
Early Myocardial Infarction Detection x List of Equations Equation (1) Fourier Transform .................................................................................................... 24 Equation (2) Short Time Fourier Transform ................................................................................. 25 Equation (3) Mother Wavelet ....................................................................................................... 26 Equation (4) Continuous Wavelet Transform ............................................................................... 26 Equation (5) Complete Equation for Continuous Wavelet Transform ......................................... 26 Equation (6)Scaling and Mother Wavelet Function of Biorthogonal Wavelet............................. 29 Equation (7)Dual Scaling and Mother Wavelet Function of Biorthogonal Wavelet .................... 30 Equation (8)Frequency Dilation for Biorthogonal Wavelet ......................................................... 30 Equation (9)Frequency Wavelet for Biorthogonal Wavelet ......................................................... 30 Equation (10)Biorthogonal Wavelet Decomposition.................................................................... 30 Equation (11)Signal Filtering Equation ........................................................................................ 41
Early Myocardial Infarction Detection 1
I.
Objective
The objective of this project is to create a smart algorithm that can detect the elements of an electrocardiogram (ECG) and determine if symptoms of myocardial infarction are present. The algorithm is written in MATLAB, but when coupled with a portable ECG machine, can provide greater protection against mortality and morbidity associated to myocardial infarction.
II.
Introduction
Myocardial Infarction (MI) is commonly referred to as “Heart Attack”. A “heart attack” is defined by World Health Federation as a condition which “occurs when the heart’s supply of blood is stopped” (World Health Federation, 2009). It is highly important to understand the meaning of the words myocardial infarction in order to diagnose the disease. The word ‘myocardial’ means related to the heart muscle and the word ‘infarction’ means tissue death due to lack of oxygen and ‘myocardial infarction’ means heart muscle or tissue death due to lack of oxygen. When the heart’s blood supply is restricted, “a sequence of injurious events occur beginning with subendocardial or transmural ischemia, followed by necrosis, and eventual fibrosis (scarring) if the supply isn’t restored in an appropriate period of time” (Yanowitz, 2006). The American heart association defines myocardial infarction as “the damaging or death of an area of the heart muscle (myocardium) resulting from blocked blood supply to the area; medical term for a heart attack” (American, 2008). The need for early diagnosis of myocardial infarction is apparent from the statistics estimated by American heart association. The United States American Heart Association estimates 80,700,000 people suffered with some form of cardiovascular symptom, out of which 8,100,000 suffered from myocardial infarction alone. The heart disease and stroke statistics for
Early Myocardial Infarction Detection 2 the year 2008 by the American Heart Association publishes that there are 600,000 new incidences of myocardial infarction reported annually and 320,000 recurrent attacks annually. Hospital stays in 2005 were recorded as 1.8 million in-patients, which amounted to $256 billion dollars in direct and indirect cost of myocardial infarction. American Heart Association’s statistics also show that Heart Attacks are still the leading cause of death in the United States of America. Myocardial Infarction is not a fatal condition if proper medical help is received at the right time. MI can be diagnosed by various diagnostic tools like Angiogram, Echocardiogram, Blood Analysis, Chest X-ray and the oldest and most trusted tool by the doctors ECG or electrocardiography. Not only is ECG the oldest tool available to monitor the electrical activity of the heart, it is also the most efficient diagnostic tool, giving speedy diagnosis compared to the other available tool for monitoring heart activity. Early recognition of symptoms of myocardial infarction can reduce the morbidity and mortality of patients. The literature shows that continuous monitoring of heart electrical activity decreases the changes of fatal myocardial infarction and the project aims at developing an algorithm that can easily be incorporated in the current available portable and wireless ECG machines to create a ‘standalone’ state of the art smart medical device for giving an early warning against the imminent myocardial infarction.
III. Anatomy of Heart It is important to understand the anatomy of the heart in order to understand what myocardial infarction is and why does it occur. Heart is a muscular organ that supplies blood through the body. It is located between the lungs in the left side of the sternum. The heart has
Early Myocardial Infarction Detection 3 four chambers as can be seen in Figure 1 below. The four chambers are Right Atrium, Right Ventricle, Left Atrium and Left Ventricle.
Figure 1: Anatomy of Heart [Source: Heart Information Center (2006)] •
Right Atrium: This chamber consists of de-oxygenated blood that returns from the body, this de-oxygenated blood is then passed on to the Right Ventricle through the tricuspid valve.
•
Tricuspid Valve: It is a one-way valve that controls the flow of blood from the Right Atrium to the Right Ventricle.
•
Right Ventricle: It is a chamber that consists of de-oxygenated blood which is passed into the lungs for oxygenation via the pulmonary valve.
•
Pulmonary Valve: It is a one-way valve that controls the flow of blood from Right Ventricle to the pulmonary arteries.
Early Myocardial Infarction Detection 4 •
Pulmonary Arteries: These arteries supply de-oxygenated blood to the lungs where the blood gets oxygenated.
•
Pulmonary Veins: After the blood passes to the lungs from the pulmonary arteries, the blood gets oxygenated and flows from the lungs to the pulmonary veins, the pulmonary veins supply the oxygenated blood from the lungs to the Left Atrium.
•
Left Atrium: This is the chamber where the oxygenated blood enters from the pulmonary vein. The blood from the left atrium is then forced into the left ventricle via the mitral valve.
•
Mitral Valve: It is a one-way valve that controls the flow of blood from the left atrium to the left ventricle.
•
Left Ventricle: The oxygenated blood enters the left ventricle through the mitral valve and is then forced from the left ventricle into the aorta through the aortic valve.
•
Aortic Valve: It is a one-way valve that controls the flow of blood from the left ventricle to the aorta.
•
Aorta: It is largest artery in the body and the aorta branches into smaller arteries. The aorta carries the oxygenated blood from the heart to the other parts of the body. The Texas Heart Institute gives some interesting and fun facts about heart and they are
listed below: •
Heart Weighs between 7 and 15 ounces (200 to 425 grams) and is little larger than the size of the fist
•
In a lifetime, the heart expands and contracts 3.5 billion times
Early Myocardial Infarction Detection 5 •
In a day, heart pumps 2,000 gallons (7,571 liters) of blood and the heart beats 100,000 times
IV. Physiology of Heart With understanding the anatomy of heart, this section would discuss the physiology of the heart, the electrical conductivity that drives the heart and pumps the blood throughout the body. The heart is made of cardiac muscle tissue that contracts and relaxes throughout the lifetime of a person and this contraction and relaxation of the muscles drives the blood from the heart. The contraction and relaxation of the cardiac muscle is in a rhythm, when the cardiac muscles of the heart’s ventricles contract, it is called as systole and when the cardiac muscles of heart’s ventricles relax, it is called as diastole. “A network of nerve fibers coordinates the contraction and relaxation of the cardiac muscles tissue to obtain an efficient, wave-like pumping action of the heart” (Cardiovascular Consultants, 2006). Figure 2 shows the diagram of heart with some of the key elements of labeled that are necessary in understanding the physiology of heart. Sinoatrial node commonly known as the SA node is the natural pacemaker of the heart. It triggers an electrical impulse that produces a heartbeat. The impulse trigger passes through the atria and causes the muscles to contract. The impulse that travels from the SA node reaches the Atrioventricular node commonly known as the AV node after the contraction of the atrium muscles. The AV node triggers another pulse which now causes the ventricles to contract.
Early Myocardial Infarction Detection 6
Figure 2: Electrical Conduction System of the Heart ‘Bundle Branch Block’ of Heart [Source: Heart Information Center (2006)] The ventricular contraction is brought about by the bundle of His which receives the triggering impulse from the AV node. The bundle of His then divides the impulse into the left bundle branch and the right bundle branch, which in turn contracts the left and right ventricles. The contraction and the relaxation of the heart muscles thus brought about by the SA and the AV nodes is wavelike and in rhythm. The rhythmic wavelike activity can be heard by the doctors using the stethoscope. It can also be imaged using echocardiography that uses the principle of ultrasound or heart imaging. Electrocardiography is another diagnostic tool for monitoring the rhythmic electrical activity of the heart. Subsequent sections will introduce the principle behind electrocardiography along with its advantages and disadvantages. This rhythmic electrical activity of the heart sometimes is lost and “the electrical impulse cannot travel throughout the heart because part of the heart’s conduction system is ‘blocked’” (Heart Information Center, 2006) due build of plaque, cholesterol deposits in the arteries that supply blood to the heart. This is one of the reasons that lead to the arrhythmic electrical activity of the heart. There are several ways to diagnose the cause of loss of rhythm in the conduction of
Early Myocardial Infarction Detection 7 heart. Chest X-ray, Angiogram, echocardiogram and electrocardiogram are some of the diagnostic tools that aid in defining the cause of blockage of the arteries supplying blood to the heart. Myocardial Infarction is one such condition that results due to the blockage of the artery supplying blood to the heart. What is myocardial infarction and how it is caused is discussed in the next section.
V.
Myocardial Infarction
Myocardial Infarction is a type of ischemic heart disease. “Myocardial infarction (MI) is the irreversible necrosis of heart muscle secondary to prolonged ischemia” (Samer Garas et al. 2008). It is caused due to relative insufficiency of oxygen to the heart muscles called cardiac muscles. Myocardial Infarction is associated with acute coronary syndrome and “approximately 90% of MIs result from an acute thrombus that obstructs an atherosclerotic coronary artery” (Samer Garas et al. 2008). Myocardial Infarction can result due to the following causes: •
“Occlusive intracoronary thrombus - a thrombus overlying an ulcerated or fissured stenotic plaque causes 90% of transmural acute myocardial infarctions” (Klatt E.C, 2008).
•
“Vasospasm - with or without coronary atherosclerosis and possible association with platelet aggregation” (Klatt E.C, 2008).
•
“Emboli - from left sided mural thrombosis, vegetative endocarditis, or paradoxic emboli from the right side of heart through a patent foramen ovale” (Klatt E.C, 2008). Narrowing and hardening of heart muscles is process that is known in the medical terms
as ‘Atherosclerosis’. Atherosclerosis when happens in the arteries that supply blood to the heart,
Early Myocardial Infarction Detection 8 it results in coronary heart diseases. There are various coronary heart diseases and myocardial infarction is one of the types of coronary heart disease. When the blood supply of the heart muscles is hampered, it can lead to chest pain called angina and if angina is not treated, it may result in heart attack. The blood vessels that supply blood to heart are called the coronary arteries. There are three coronary arteries that supply blood to the heart. These three coronary arteries supply blood to three different areas of heart muscles cells. Since the area of these heart muscles cells is known, myocardial infarction “can occur in the anterior, lower, and the lateral heart territory” (The Myocardial Infarction-Heart Attack, 2006). Over the years cholesterol and other fatty substances in the blood get deposited on the arterial wall and builds to form a ‘Plaque’ or ‘Atheroma’. The plaque build up obstructs the flow of blood to the heart muscle. This is depicted in the figure 3 below.
Figure 3: Myocardial Infarction [Source: Coronary Artery Disease, 2008] The obstruction of blood flow across the coronary artery develops a pain in the chest called angina. It also develops pain in the arms, around the neck and back of the chest. Necrosis
Early Myocardial Infarction Detection 9 of heart tissue begins when the plaque bursts and causes a blood clot that blocks the artery completely. This cuts off the blood supply to an area of the heart. As the supply of oxygen to the blocked artery stops, the heart muscle cells start to die. The dying heart muscle cells constitute the zone of infarction as can be seen in figure 4. The area of heart muscle cells surrounding the zone of infarction is called the zone of injury (as seen in figure 4), the heart muscle cells in this area do not die but the working of those heart cells is hampered to an extent that the cells are rendered non-functional. This is called a heart attack or myocardial infarction; it causes permanent damage to that area of the heart. The zone surrounding the zone of injury is the zone of ischemia; the heart muscle cells in this region are partially functional. If enough heart muscle is damaged the heart may beat irregular or may even stop beating altogether.
Figure 4: Resulting zones from Myocardial Infarction [Source: Myocardial Infarction (2009)] There are patterns that are observed by the cardiologists in prognosis of myocardial infarction cases, they are as follows: •
“Transmural infarct - involving the entire thickness of the left ventricular wall from endocardium to epicardium, usually the anterior free wall and posterior free wall and
Early Myocardial Infarction Detection 10 septum with extension into the RV wall in 15-30%. Isolated infarcts of RV and right atrium are extremely rare” (Klatt E.C, 2008). •
“Subendocardial infarct - multifocal areas of necrosis confined to the inner 1/3-1/2 of the left ventricular wall. These do not show the same evolution of changes seen in a transmural MI” (Klatt E.C, 2008). When myocardial infarction occurs there is gradual necrosis of heart tissue. The necrosis
of the heart tissue happens over a period of time and may vary with depending on the size of the infarct. Table 1 gives the pathologic findings in terms of timeline for the necrosis of heart tissue. It is well known that first myocardial infarction may not be fatal but subsequent infarcts may prove fatal depending on the damage done by the previous infarcts. This also would change the timeline for the necrosis of heart tissue and may result in sudden death. “Sudden death is defined as death occurring within an hour of onset of symptoms” (Klatt E.C, 2008). Some preexisting conditions that may prove dangerous for subsequent infarcts are as follows: •
“Arrhythmias and conduction defects, with possible ‘sudden death’” (Klatt E.C, 2008).
•
“Extension of infarction, or re-infarction” (Klatt E.C, 2008).
•
“Congestive heart failure (pulmonary edema)” (Klatt E.C, 2008).
•
“Cardiogenic shock” (Klatt E.C, 2008).
•
“Pericarditis” (Klatt E.C, 2008).
•
“Mural thrombosis, with possible embolization” (Klatt E.C, 2008).
•
“Myocardial wall rupture, with possible tamponade” (Klatt E.C, 2008).
•
“Papillary muscle ruptures, with possible valvular insufficiency” (Klatt E.C, 2008).
•
“Ventricular aneurysm formation” (Klatt E.C, 2008).
Early Myocardial Infarction Detection 11 After understanding what and how myocardial infarction occurs, it is now essential to know how to get myocardial infarction diagnosed. As mentioned already in the introduction there are several ways for diagnosis of myocardial infarction. The next section will describe a few different diagnostic tools that diagnose myocardial infarction. a. Diagnostic Studies to detect Myocardial Infarction The diagnostic tools for studying myocardial infarction have been divided into three types and they are as follows: i. Blood Analysis During the period of myocardial infarction the dying heart muscle cells release different types of enzymes called as ‘cardiac enzymes’. These different cardiac enzymes are present in the bloodstream at different intervals of time. The cardiac enzymes can be seen in the bloodstream as early as start of the infarct. Blood analysis for these different enzymes can reveal the crucial information for diagnosis of myocardial infarction. The different cardiac enzymes are as follows: •
“Troponin” (Samer Garas et al.,2008) This enzyme plays a major role in diagnosis of myocardial infarction. The reason is
because “Troponin I and T are structural components of cardiac muscle” (Klatt E.C., 2008). The level of troponins can be found in the bloodstream as early as 3 to 12 hours in myocardial infarction. The levels of this enzyme will also remain elevated in the bloodstream for up to 2 weeks into myocardial infarction. The reason for troponin not being considered as the only blood analysis enzyme is because another account of myocardial infarction happening in the time period of already elevated levels of troponins will go unnoticed and may prove fatal.
Early Myocardial Infarction Detection 12 Table 1: Macroscopic & Microscopic Findings of MI [Source: Klatt.E.C, 2008] Time of Onset 18-24 Hours 24-72 Hours 3-7 Days 10-21 Days 7 Weeks
Gross Morphologic Findings Pallor of myocardium Pallor with some hyperemia Hyperemic border with central yellowing Maximally yellow and soft with vascular margins White fibrosis
Time of Onset 1-3 Hours 2-3 Hours 4-12 Hours
Microscopic Morphologic Finding Wavy myocardial fibers Staining defect with tetrazolium or basic fuchsin dye Coagulation necrosis with loss of cross striations, contraction bands, edema, hemorrhage, and early neutrophlic infiltrate
18-24 Hours
Continuing coagulation necrosis, pyknosis of nuclei, and marginal contraction bands
24-72 Hours
3-7 Days
Total loss of nuclei and striations along with heavy nuetrophilic infiltrate Macrophage and monoclear infiltration begin, fibrovascular response begins
10-21 Days 7 Weeks
Fibrovascular response with prominent granulation Fibrosis
Early Myocardial Infarction Detection 13 Also, troponin elevations have been found in other forms of conditions like renal failure and also conditions related to skeletal muscles. •
“Creatine Kinase” (Samer Garas et al.,2008) This enzyme consists of 3 sub-enzymes. Each type of enzyme is related to a particular
part of the body. “creatine kinase with muscle subunits (CK-MM), which is found mainly in skeletal muscle; creatine kinase with brain subunits (CK-BB), predominantly found in the brain; and myocardial muscle creatine kinase (CK-MB), which is found mainly in the heart” (Samer Garas et al.,2008). Creatine kinase (CK-MB) along with troponin is usually obtained together to provide better diagnosis of myocardial infarction. The sensitivity of creatine kinase is approximately 95%. •
“Myglobin” (Samer Garas et al.,2008) This is a type of protein that is found in both skeletal as well as cardiac muscles. The
function of myoglobin is to bind oxygen. This makes the identifying level of myoglobin in the bloodstream important as it would help in determining the amount of injury made to the heart muscle in myocardial infarction. Rise in the level of myoglobin is present even before creatine kinase-MB, but the rise may or may not be related to myocardial infarction alone. •
“Lactate Dehydrogenase” (Klatt E.C., 2008) This enzyme like creatine kinase consists of 5 different enzymes and like all the other
enzymes does not help in diagnosis of myocardial infarction alone. When used in conjunction with the other types of enzymes, the blood analysis for these enzymes gives excellent myocardial
Early Myocardial Infarction Detection 14 infarction diagnosis. “It begins to rise in 12 to 24 hours following MI, and peaks in 2 to 3 days, gradually dissipating in 5 to 14 days” (Klatt E.C., 2008). There are other blood analysis tests that are done as part of diagnosis for myocardial infarction and they are as follows: •
Complete blood cell count
•
Chemistry Profile
•
Lipid level Profile
•
C-reactive Protein (CRP)
These above mentioned blood analysis types are not specifically related in diagnosis of myocardial infarction alone and are utilized as diagnostic methods for other types of conditions and diseases as well. ii. Imaging There are several Imaging diagnostic tools available for diagnosis of myocardial infarction. The imaging diagnostic tools are as follows: •
“Chest Radiography” (Samer Garas et al.,2008) Chest Radiography is another word for Chest X-ray. This imaging technique does not
provide results that are specific to myocardial infarction detection, but is usually done as one of the first step towards assessing any patient admitted in an emergency room with a heart condition. A chest Radiograph is used for assessing the size of the heart and also conditions such as pneumonia that might be one of the reasons for certain types of heart conditions. The details
Early Myocardial Infarction Detection 15 revealed in a chest radiogram are mostly anatomic and macroscopic, for microscopic details other imaging tools like angiogram proves helpful. •
“Echocardiography” (Samer Garas et al.,2008) Echocardiogram is an excellent imaging tool for cardiologists. It provides “the extent of
the infarction and assesses overall left ventricle (LV) and right ventricle (RV) function” (Samer Garas et al., 2008). It also helps in diagnosis of different complications of the heart valves. Echocardiography is an imaging tool is generally required by the physicians when all the other tests are questionable or inconclusive. •
“Myocardial Perfusion Imaging” (Samer Garas et al.,2008) This imaging tool is generally used for patients with after heart attack to assess the
damage done to the heart muscle tissue by previous infarctions. This tool is also used for prognosis of patients entering the emergency room with serious heart conditions. •
“Cardiac Angiography” (Samer Garas et al.,2008) Cardiac Angiography more commonly known as angiogram is a technique where a radio
opaque dye is inserted in the blood stream via the femoral artery and then a chest X-ray of the patient is obtained to diagnose blockages in the coronary arteries. This type of imaging technique has become very common in the recent years. The process through which is radio opaque dye is inserted is called cardiac catherization. It is an imaging technique which is minimally invasive and requires local anesthesia.
Early Myocardial Infarction Detection 16 iii. Electrical Activity Monitoring Probably the most reliable and oldest available tool for measuring electrical activity of the heart is Electrocardiography or more commonly known as ECG. ECG is considered as the first diagnostic tool when evaluating patients with suspected Myocardial Infarction. It is confirmatory of the diagnosis in approximately 80% of cases. Electrocardiography when used as a diagnostic tool in myocardial infarction can yield the following prognosis: •
To rule out the Right Ventricular infarct, ECG recording on the right-sided setting for patients with inferior Myocardial Infarction (Samer Garas et al., 2008).
•
Obtain daily serial ECGs for the first 2-3 days and additionally as needed (Samer Garas et al., 2008).
•
Convex ST-segment elevation with upright or inverted T waves is generally indicative of MI in the appropriate clinical setting (Samer Garas et al., 2008).
•
ST depression and T-wave changes may also indicate evolution of NSTEMI (Samer Garas et al., 2008). A more detailed description of what electrocardiography is and how it is used as a
diagnostic tool for monitoring electrical activity of the heart is given in the following section.
VI. Electrocardiography Patients suffering from S-T elevated myocardial infarction can be diagnosed with several type of diagnostic methodology. One of which is by using electrocardiography. The recorded trace of electrocardiography is called an electrocardiogram (ECG).
Early Myocardial Infarction Detection 17 Electrocardiography is a non-invasive diagnostic procedure that records the electrical current transmitted by the heart all over the body. The electrical current can be picked up by an electrical sensing device, which is attached to an ECG machine. The electrical sensing devices are strategically placed on the body surface to detect heart impulses. They can be placed in the arms and legs. The recorded ECG is the representation of the depolarization and re-polarization of the heart and can diagnose a patient by looking at the characteristics of the traced ECG readings (Klabunde, R.E., 2007). Every beating of the heart, an electrical impulse is generated and transmitted to the myocardium which causes the pumping action of the heart and provides blood throughout the body. There are 3 main deflections in an ECG: the P-wave, the QRS complex, and the T-wave. The P-wave records the electrical activity of the atria. It starts from the SA node, which generates an impulse. It causes an excitation to the atria and I then picked up by the AV node. The P-wave usually lasts for about 0.8 to 0.1 seconds (80 to 100 ms). The impulse travels from the AV node to the Bundle of His, which generates an isoelectric pattern in the ECG. A trace between the onset of the P-wave to the onset of the QRS complex is called the P-R interval. It is the representation of the onset atrial depolarization and the onset of ventricular depolarization. The P-R interval has a period of 0.12 to 0.20 seconds (120 to 200 ms). The next main deflection of an ECG is the QRS complex. It represents the ventricular depolarization. The impulse travels from the Bundle of His to the ventricular walls through the left and right bundle branches. The impulse causes contraction of the ventricular walls, which causes the blood to be pumped out to the lungs and body. The QRS complex has a short
Early Myocardial Infarction Detection 18 duration, usually lasts for 0.06 to 0.1 seconds (60 to 100 ms). The short duration indicates that the ventricular depolarization happens in a relatively quick time. After the QRS complex, an isoelectric line called an S-T segment occurs, at which, there is a complete depolarization in the ventricles. The isoelectric line of the S-T segment is very important in diagnosing heart conditions in a patient, since a depressed or elevated S-T segment represents a cardiac ischemia. The last major deflection in an ECG is the T-wave. The T-wave represents the repolarization of the ventricles. It has a longer trace in the ECG reading. It is sometime followed by a U-wave, which represents the remainder of the ventricular re-polarization. The trace between the ventricular depolarization and re-polarization is called the Q-T interval. The range of the Q-T interval is between 0.2 and 0.4 seconds (200 and 400 ms). The duration of the Q-T interval is important in detecting certain types of tachyarrhythmias. a. Measuring ECG Electrical sensing devices or electrodes are placed strategically on top of the body to detect the electrical activity of the heart and diagnose patients with different heart anomalies. In addition, the recorded trace of the ECG is not always recorded as shown in figure 5. The trace depends on the position of the lead. The leads are placed on the body can be described as a positive lead and a negative lead. The electrical impulse that is generated in the heart travels in parallel to the direction of the lead. If the direction moves toward the positive lead, then a positive deflection takes place, on the other hand, if the direction of the impulse moves toward the negative lead, then a negative deflection takes place. This action is illustrated in figure 6.
Early Myocardial Infarction Detection 19
Figure 5: An ECG with the major peaks and intervals. (Vibes Electrocardiogram, n.d.)
Figure 6: Illustrates the cause of deflection of an ECG (O’ Grady, M.R., n.d.)
Early Myocardial Infarction Detection 20 Electrodes are placed in the arms and legs, which are called the Einthoven's triangle. The Einthoven’s triangle, shown in figure 7, is composed of Leads I, II, and III, which are the basic electrodes of the 12 lead ECG system.
_
+ Lead I
_
_
Lead II
Lead III
+
+
Figure 7: An Einthoven's triangle with Lead I, II, and III. There are 3 additional leads that are developed from leads I, II, and III. They are called the augmented limb leads, aVR, aVL, and aVF. The augmented limb leads views the heart in different vectors compared to the original leads. It is the recording between one limb and two other limbs. aVR, augmented vector right, is a recording between the positive lead in the right arm and a combination of negative leads in the left arm and left leg. aVL, augmented vector left, is between the positive lead of the left arm and combination of negative leads in the right arm and left leg. aVF, augmented vector foot, is between the positive lead of the left foot and combination of right arm and left arm. Leads I, II, III, aVR, aVL, and aVF represents the hexial reference system which determines the electrical axis of the heart. All leads are in the frontal plane axis as shown in figure 8.
Early Myocardial Infarction Detection 21
Figure 8: Axial representation of Lead I, II, III, aVR, aVL, and aVF. (O'Grady, M.R., n.d.) Precordial leads, V1, V2, V3, V4, V5, and V6, are also called chest leads because they are strategically placed on the chest overlying the heart or its vicinity. Unlike leads I, II, III, aVR, aVL and aVF, the precordial leads records ECG in the horizontal plane. The precordial leads are V1, V2, and V3 are called thee right precordial leads; while V4, V5, and V6 are called left precordial leads and all precordial leads are unipolar, which means it can be a positive lead or a negative lead. The precordial leads provide different view of the electrical activity of the heart and they are very useful in identifying the P P-wave of an ECG recording. The ECG waveform follows a pattern that describes the physiological meaning of the conduction of heart. Figure 9 shows such a typical waveform. The description for the typical waveform veform with its analogous conduction activity is as follows:
Early Myocardial Infarction Detection 22 •
“The first deflection, termed the P-wave is due to the depolarization of the atria” (Jouck. P.P.H., 2004).
•
“The large QRS-complex is due to the depolarization of the ventricles. This is the complex with the highest amplitude” (Jouck. P.P.H., 2004).
•
“The last and most significant part for this report is the T-wave. It corresponds to the ventricular repolarization of the heart” (Jouck. P.P.H., 2004).
Figure 9: Typical ECG waveform [Source: Jouck. P.P.H. (2004)]
Early Myocardial Infarction Detection 23 VII. Wavelet Transforms Signal transformation aides in converting signal from Time domain to Frequency domain. It is of very importance to be able convert the signal from time domain to frequency domain to get the complete information carried by the signal. There are many different mathematical transformations available for converting time domain signal to frequency domain and vice a versa. Typically physiological signals are present in time domain i.e. the signal has a time value and the amplitude value. What is also hidden in this type of signal is its frequency value. When the time domain signal is converted in frequency domain, the hidden frequency values of the signal can be found out. It is important to know the frequency component of a signal to completely define and/or analyze the signal. Hence, mathematical signal transformation plays a crucial role in signal analysis and signal representation that are the key elements in determining the ECG signal and hence important for this project. There are several mathematical transformation tools available, hence choosing the transformation tool that best suites the signal that is under measurement is of high importance. Some of the mathematical signal transformation tools are listed below: •
Fourier Transform
•
Short Time Fourier Transform
•
Wigner Distribution
•
Laplace Transform
•
Z-Transform
•
Wavelet Transform
•
Fast Fourier Transform
•
Hilbert Transform
Early Myocardial Infarction Detection 24 •
Radon Transform
•
Linear Canonical Transform The list of transform above is no way a complete list; there are several other
mathematical transformations available. Choice of transform for a particular type of application is critical. There are certain conditions based on which the choice of the transform to be used for a particular application is based on. The conditions were crucial in deciding type of transform to be used for ECG signal processing were as follows: •
Type of application
•
Type of signal, stationary or non-stationary The choice of transform made for this project was Biorthogonal Wavelet Transform. The
reason for choosing this type of mathematical signal transformation was done methodologically. For most of the engineering applications, Fourier transform is popular (Polikar Robi, 2001). Fourier transforms can provide signal representation either in time domain or the frequency domain. The equation 1 for Fourier Transform is given as follows:
………………………..(1)
What Fourier transforms fails to provide is the time instant at which the frequency value is present and hence is not a transform of choice for non-stationary signal. Since ECG signal is non-stationary signal Fourier transform was not a favored choice for doing ECG signal analysis in this project. The next suitable mathematical transformation is the Short Time Fourier Transformation. As the name suggests, short time Fourier transform is similar to Fourier transform but there is a
Early Myocardial Infarction Detection 25 small difference between then which is “In STFT, the signal is divided into small enough segments, where these segments (portions) of the signal can be assumed to be stationary” (Polikar Robi, 2001). The equation 2 for Short Time Fourier Transform is given as follows: #
# ′
!"#
$……………… (2)
Short Time Fourier Transform does overcome the limitation of signal representation with known time instant of frequency. But there is a limitation to using Short Time Fourier Transform too, which is the segments of the signal called window have fixed width and it may not be useful for a signal that has multiple frequencies varying rapidly throughout the signal. If the width of the segment or the window is chosen such that it narrow, then the signal transformation obtained has good time resolution but poor frequency resolution, on the other hand, if the width of the window is chosen such that it is wide, then the signal transformation obtained has poor time resolution but good frequency resolution. For getting both good time resolution as well as good frequency resolution, the width of the window should be varied across the signal during the signal transformation such that both time and frequency component of the signal does not get compromised. Due to this limitation of Short Time Fourier Transform, Wavelet Transform was originally developed. For this project, the signal under measurement ECG that is non-stationary signal that has varying frequency levels throughout the signal, and with addition of noise, it would be difficult to use short time Fourier transform for this project, hence Wavelet Transform was chosen. Like short time Fourier transform, wavelet transform are similar “in the sense that the signal is multiplied with a function, similar to the window function in the STFT, and the transform is computed separately for different segments of the time-domain signal” (Polikar
Early Myocardial Infarction Detection 26 Robi, 2001). There are differences in Short Time Fourier Transform and Wavelet Transform though and they are as follows: •
“The Fourier transforms of the windowed signals are not taken, and therefore single peak will be seen corresponding to a sinusoid, i.e., negative frequencies are not computed” (Polikar Robi, 2001).
•
“The width of the window is changed as the transform is computed for every single spectral component, which is probably the most significant characteristic of the wavelet transform” (Polikar Robi, 2001). The continuous wavelet transform has a time scaling and a time shifting function ψ (t)
called as the mother wavelet. The time scaling and the time shifting function given by equation 3 and the continuous wavelet transform are given by equation 4 and 5. ψ%τ &
'
()%)
ψ*
+ τ %
,-……………………………………(3)
.- / -0-1-
Where
2& $ 3 )56)7 -$6 )6)
45
8 -9
“The conditions above state that ψ (t) is bandpass and sufficiently smooth. Assuming that || ψ (t) || = 1, the definition above ensures that ||2%: &||=1 for all .-and 0” ( Schniter Phil,2005).
CCCCCCCCC ; ? @ >A B -……………………………..(4)
D ;?--------@>A B ?>…………………(5) > ;@
Early Myocardial Infarction Detection 27 Explaining the above equations in simple words mean that continuous wavelet transform can decompose the signal into a collection of shifted and stretched versions different scales. Wavelet transform is thus a ‘time-scale’ analysis and scale is inverse of frequency. There are different types of wavelets available and few of the wavelet transforms are listed below: •
Haar
•
Morlet
•
Mexican Hat
•
Meyer
•
Quadratic Spline
•
Dyadic Spline Wavelet
•
Debauchies
•
Biorthogonal
•
Gaussian There are several other types of wavelets and each type of wavelet is different from one
to the other by properties such as symmetry, singularity etc. The selection of wavelets for ECG signal processing is done based on the following parameters: •
“Orthogonal Vs. Non-Orthogonal: a non-orthogonal analysis involves high redundancy at larger scales” (Bhatia et al, 2006).
•
“Complex or real valued wavelet function: a complex wavelet providing information about both amplitude and phase is better suitable for oscillatory signal behaviour whereas real valued wavelet function only returns a single signal
Early Myocardial Infarction Detection 28 modulus that can be used to isolate signal peaks and discontinuities” (Bhatia et al, 2006). •
“Width of the wavelet function: this directly acts on the analysis resolution that is for wavelet the result of balance between the length of analysis of samples window in time frequency axes” (Bhatia et al, 2006).
•
“Shape of the mother wavelet: wavelet filtering can be viewed as an adapted filter looking for the highest correlation between the ECG signal to analyze and the considered wavelet” (Bhatia et al, 2006).
“For ECG parameter estimation, it is desirable that the basis function (wavelets) be symmetric/antisymmetric” (Sivannarayana et al. 1999). “It is also desirable that the basis have a minimum number of sign changes which will simplify the steps in the parameters estimation algorithm” (Sivannarayana et al. 1999). Out of the wavelets listed above Biorthogonal Wavelet was the choice of wavelet for this project because it satisfies the properties that are required for ECG signal analysis. To understand Biorthogonal Wavelets, it is important to understand the concepts of dual basis. “Consider a two-dimensional coordinate space. Any two vectors {e1, e2} that are not parallel can form a basis for the space. If the angle between the two vectors happens to be 90 degrees, we have an orthogonal basis” (Wolfram Research, 2009). “Any vector F in this space can be written uniquely as a superposition of the two basis E
vectors: F G' ' H G . If the basis is orthogonal, I JI- and the component GI along I is E
given by the inner product I F G' I ' H G I GI . However, if the basis is not E
orthogonal, GI is no longer given by the inner product of Fand I . In order to calculate the E
componentGI , we introduce another set of basis vectorsK 'L L M, called the dual of {e1, e2}. The
Early Myocardial Infarction Detection 29 dual basis satisfies IL N JI- and the space spanned by the dual basis is called the dual space of the original space. In terms of the dual basis, the components of a vector along the basis {e1, e2} can be calculated as IL NF - O G IL N GI ”(Wolfram Research, 2009) . “Similarly, for a E
nonorthogonal basis KPI M of a function space, we can introduce dual basis KPIL M L CCCCCCCC byQPIL P R - ∞ P S -P $ - JI- . A function f(t) can be decomposed as a superposition ∞
of the nonorthogonal basisKPI M using a set of dual basis functionsKPIL M: OI I PI OIPIL PI . We will tacitly assume that the function space and its dual are the same, a condition satisfied by L2. Therefore, the roles of dual basis and the original basis can be interchanged and we obtain OIPI -PIL . When KPI M is orthogonal, we have the obvious relation P L P” (Wolfram Research, 2009). “The dilations and translations of the scaling function KP TM constitute a basis for Vj and, similarly, KU TM for Wj. To define a dual multiresolution analysis with dual subspacesKV L M and KW L M generated from a dual scaling function P L and a dual mother waveletU L , respectively. In terms of subspaces, the above biorthogonality conditions can be expressed as VX Y - W L V L Y WX-.Z$-WX Y W L [\-X ] X L ” (Wolfram Research, 2009). “By definition, a scaling function and a mother wavelet satisfy the dilation equation and the wavelet equation, respectively. So we have ^ - _ Oa `a ^ a>b-c _ Oa da ^ a-…………………………(6) and
Early Myocardial Infarction Detection 30 L ^L - _ Oa `aL ^L a>b-cL _ Oa dL a ^ a-…………………..(7)
The roles of the two functions-P andP L or U and-U L can be interchanged. Or if we "take the dual" of the above equations (define .e .), we obtain the following relations, L f
CCCCCCCCCCCCC QP'f P L R - _g Pg TP L $-.Z$-hfL - QP'f U L R - _g Pg
T-U L $” (Wolfram Research, 2009). “In frequency space, the above dilation and wavelet equations 6 and 7 assume the form j
j
j
j
^i j k- * , ^i * , ->b-ci j l- * , ^i * ,- ……………………………….(8)
and j
j
j
j
^Li j k- * , ^Li * , ->b-ci j l- * , ^Li * ,-- …………………………..(9)
As before,m is defined by-m - Of
f
Ifn_
, and G, m L , ando L are
defined analogously” (Wolfram Research, 2009).
The “biorthogonality conditions can be translated into conditions on the filter coefficients using the dilation and wavelet equations. These conditions are CCCCCCCCCCCCC CCCCCCCCCCCC CCCCCCC H m L H pm CCCCCCC H o L H po m L m H p q-.Z$-o L o H p q-CCCCCCCCCCCCC CCCCCCCCCCCC CCCCCCC H m L H po CCCCCCC H o L H pm H p 3-.Z$--m L o H p 3-o L m Using P r T Os
s
gTPXr and U r T Os hs gTPXr, the relations for
wavelet decomposition becomes” (Wolfram Research, 2009):
D
a
D
Ot `t a-t >b-a
Ot dt a-t -…………………………………….(10)
Early Myocardial Infarction Detection 31 Figure 11 below shows the typical shapes of biorthogonal and reverse biorthogonal wavelets available in the Matlab 7.1 Release 14 Wavelet Toolbox 3.0 Biorthogonal Wavelets 2.4 were chosen because of the near shape of the ECG signal and that of the 2.4 biorthogonal wavelet. This can be seen in the figure 10.
Figure 10: 2.4 Biorthogonal Wavelet and ECG Signal
Early Myocardial Infarction Detection 32
Figure 11: Types of Biorthogonal Wavelets available in Wavelet Toolbox 3.0 of Matlab 7.1 R 14[Source: Matlab7.1R14 Help]
Early Myocardial Infarction Detection 33 VIII. Introduction to Early Myocardial Infarction Detection System After having acquired the background for understanding how myocardial infarction occurs and what the ways and means to diagnose it are, we move on to discuss about the specifics that are crucial to understand the algorithm for early myocardial infarction detection. The subsequent section will discuss the specifications that were considered, the features of the algorithm, the detailed functions, and the steps to get to early myocardial infarction detection. During the last few years telemetry monitoring has become the most widely used for of monitoring system and telemetry monitoring of cardiovascular diseases is gaining popularity. Hence, it is important to have a telemetry device monitoring the condition of heart to warn onset of myocardial infarction. The method presented in the project is to detect ST-changes in the ECG of the patient based on wavelet signal processing technique to warn onset of myocardial infarction. The testing for the method is done using the following databases from Physiobank organization: •
MIT-BIH ST change Database
•
Long-term ST Change Database
•
European ST-T Database
•
MIT-BIH Normal Sinus Rhythm Database
•
MIT-BIH Noise Stress Test Database a. Background The typical frequency range for ECG signal is between 0.5 to 100Hz. Table 2 below
gives the typical wave durations and amplitudes that are present in ECG signal which is the physiological signal under measurement for this project. Figure 12 shows the normal ECG
Early Myocardial Infarction Detection 34 waveform on a strip chart paper. It shows how to interpret data from a strip chart into the actual amplitude and time values that are of interest. “Changes in the ST-segment of the ECG may indicate that there is a deficiency in the blood supply to the heart muscle” (Tompkins Willis, 2000).
Figure 12: Normal ECG waveform on Strip Chart [Source: Barron Jon, 2007] The detection of the ECG waveform is based on the on the duration and amplitude measurements given in the table 2 and figure 10. Matlab 7.1 was used for the programming of the code for ST-elevation detection. The signal processing for the ECG waveform was done by using Biorthogonal Wavelets. For using Wavelets Wavelet Toolbox3.0 from the Matlab software was used. The use of Biorthogonal Wavelet was based on the symmetry, and the fact that the shape of the biorthogonal wavelets is closest to that of the ECG Waveform, giving precision accuracy to the time-scale conversion. By using Biorthogonal wavelets it is possible to get complete reconstruction of the signal. “The ST-segment represents the period of the ECG just after depolarization, the QRS complex, and just before the repolarization, the T- wave”(Tompkins, Willis, 2000). The ST-
Early Myocardial Infarction Detection 35 segment is isoelectric in normal ECG and hence the elevation of ST-segment is tested for by comparison between isoelectric line and the ST-segment. The next section describes the features and functions of early myocardial infarction detection. Table 2: Typical Amplitudes and Durations for ECG signal [Source: Saritha. et.al. (2008)] Amplitude P-Wave
0.25 mV
R-Wave
1.60 mV
Q-Wave
25% of R-Wave
T-Wave
0.1 to 0.5 mV Duration
P-R Interval
0.12 to 0.20 s
Q-T Interval
0.35 to 044 s
S-T Interval
0.05 to 0.15 s
P-Wave Interval
0.11 s
QRS Interval
0.09 s
The most important function of early myocardial infarction detection is that it warns the patient of the imminent attack as early as 20 minutes before the actual attack. In typical cases, a patient would only go to the emergency room when the heart attack has already happened. This reduces the time for the doctors to treat it. If the patient and the doctor are warned early of the imminent attack, the doctor gets the much needed time to curb the intensity of the attack by providing medication faster. By having the ST-elevation detection program installed on a
Early Myocardial Infarction Detection 36 portable ECG machine having wireless transmission capability will make the best standalone telemetry device available for treating myocardial infarction related conditions. b. Materials and Method i. Database Description As described above there were five databases that were used for testing the method proposed for detecting onset myocardial infarction. The databases are developed and managed by Physiobank organization. Physiobank is a database of “well-characterized digital recordings of physiologic signals and related data for use by the biomedical research community” (Goldberger et al. 2000). MIT-BIH ST change database has ECG recordings from long “exercise stress tests and exhibit transient ST depression” (Albrecht P., 1983). There are some recordings in this database that consists of ST-elevation too. Long-term ECG database consists of ECG recordings from 80 subjects “chosen to exhibit a variety of events of ST segment changes, including ischemic ST episodes, axis-related nonischemic ST episodes, episodes of slow ST level drift, and episodes containing mixtures of these phenomena” (Franc Jager. et al, 2003). European ST-T database consist of “ambulatory ECG recordings from 79 subjects” (Taddei, A.et al. 1992). MIT-BIH Normal Sinus Rhythm database consists of recordings from 18 subjects. The database comes from the Arrhythmia Laboratory at the Boston’s Beth Israel Deaconess Medical Center (Goldberger et al. 2000).
Early Myocardial Infarction Detection 37 The fifth and the last database that was tested for the method described in this report for ST-segment changes was that of MIT-BIH Noise Stress Test Database. This database consists of ambulatory ECG recordings. “The noise recordings were made using physically active volunteers and standard ECG recorders, leads, and electrodes; the electrodes were placed on the limbs in positions in which the subjects' ECGs were not visible” (Moody GB, 1984). Out of the available ECG recordings from the databases described above, 17 recordings were tested from MIT-BIH ST Change Database, 5 from MIT-BIH Noise Stress Test Database, 6 from the Long-Term ST Database, 13 from the MIT-BIH Normal Sinus Rhythm Database and 14 recording were randomly selected and tested for ST-segment change using the method described in the next section. ii. Method for ST-elevation detection The warning of the imminent myocardial infarction is done through the use of different filtering techniques, followed by the signal processing to ensure accurate ECG signal extraction and estimation. Electromagnetic Interference (EMI), muscle activity (EMG), bowel movements, and electric line interference are often always embedded in ECG signal and constitute noise in the ECG signal. It is important to be able to remove this noise in order to have a good ECG parameter estimation. The presence of this noise also changes the baseline for the ECG signal, it is also important for ECG parameter extraction and estimation to be able to remove the baseline wander before the actual ECG signal parameters are extracted. This is done by having a baseline wander cancellation along with efficient signal filtering to remove electric line, EMI, EMG and other types of noise from the ECG signal. After eliminating the baseline wander the ECG signal was converted into time-scale domain for further analysis. Converting the ECG signal into timescale domain was done because “Wavelet Transformation (WT) has shown to be substantially
Early Myocardial Infarction Detection 38 noise proof in ECG segmentation and thus appropriate for ST-T segment extraction” (Milosavljević Nebojša, et al.2006). The signal was decomposed into 4 scales ranging from 21 to 24. Figure 13 shows sample decomposed scales using Dyadic Wavelets for ECG signal.
Figure 13: Dyadic Wavelet Transform of ECG signal [Source: Jouck. P.P.H. (2004)]
It can be interpreted from Figure 13 that wavelet transform “at small scales reflects the high frequency components of the signal and, at large scales, the low frequency components. The energy contained at certain scales depends on the center frequency of the used wavelet” (Jouck. P.P.H., 2004).
Early Myocardial Infarction Detection 39 24 scale was used to detect the R-peak because “most energies of a typical QRS- complex are at scales 23 and 24…. QRS complex with high frequency components, the energy at scale 22 is larger than that at 23” (Jouck. P.P.H., 2004). According to Wenli Chen et al. 2007, the high frequency noises like the electric line interference, muscle activity, bowel movement activity, electromagnetic interference is concentrated in the lower scales of 21 and 22, while the levels 23 and 24 constitute for less noise compared to the lower scales. This can also be seen in figure 14. Wenli Chen et al. 2007 summarize that the frequency of the QRS complex is mainly present in the 23 and 24 scales. Since the scale 24 shows less noise compared to 23, in this project we chose scale 24 for R-peak detection.
Figure 14: Biorthogonal Wavelet Transform of ECG Signal from 21 to24 level The R-peak detection was followed by detecting point S and Q. (Pam, Tompkins, 1985) method was utilized for detection of the R-peaks. After finding R-peaks (Tompkins, 2000) method was used for detecting points S, Q, T-peak, T-point, J-point as seen in figure 15.
Early Myocardial Infarction Detection 40
Figure 15: Method for ECG Parameters Detection [Source: Tompkins, 2000] The first inflection point before Q point was chosen as point K and P point was found, the distance between point P and point K is the isoelectric line. The isoelectric line was then compared to the ST-segment for checking the elevation or depression of the ST-segment in all the ECG waveforms. The algorithm for the ST-change detection program is divided in several subsections and is as follows: 1. Signal Filtering and Baseline Wander Correction After getting the ECG dataset, the first step was to remove the inherent noise from the ECG signal. The typical noises that affect the ECG signal are electric or power line interference of 60Hz, EMG or the muscle activity that gets captured along with the ECG measuring electrodes, bowel movements also called EGG movement; EEG sometimes may interfere with ECG signal and constitute noise. Electromagnetic interference is also seen in the ECG signal. It thus becomes important to remove the noise from the signal for accurate and precise ECG
Early Myocardial Infarction Detection 41 parameter detection and extraction. The filtering technique applied in this project is a simple FIR filter.
Figure 16: Filter expressed in Direct Form II transposed structure [Source: Matlab7.1R14 Help] Figure 16 can also be expressed as: y(n)=(1)*x(n)+b(2)*x(n-1)+…+b(nb+1)*x(n-nb)-a(2)*y(n-1)-…-a(na+1)*y(n-na)….(11) After filtering the signal using the FIR filter, the filtered signal was passed through median filters to correct for the baseline wander correction. The process followed included passing the filtered signal through a 200ms median filter that eliminates the QRS complex from the signal. This median filtered signal was again passed through a 600ms median filter to eliminate the T-wave from the signal. This final filtered signal is the signal that consists of the noise that changes the baseline through the signal. The difference between the FIR filtered signal and the final median filtered signal thus gives the baseline wander eliminated signal. The baseline wander elimination process in shown in figure 17, where the first plot if of the FIR filtered signal, the second plot is of the first median filtered signal, while the third plot is of the second median filtered signal, the final plot shows the baseline wander eliminated signal.
Early Myocardial Infarction Detection 42
Figure 17: Baseline Wander elimination 2.
R-Peak Detection
The R-peak detection was carried out by using the Pan, Tompkins 1985 method of Rpeak detection. The method uses the threshold level to calculate the maximum amplitude in the ECG waveform. The threshold level set for the R-peak detection in the ST-change program is approximately 0.6. The R-peak detection was done in the time-scale domain at level 24. Same level was utilized to estimate the other key points in the ECG waveform. Figure 18 shows the final ECG waveform with all the detected points along with its legend. 3.
Heart Rate Measurement
It is essential to calculate the heart rate of the patient in order to determine accurate measurement and approximation of the PQSTJK points on the ECG waveform; hence heart rate was calculated from the datasets after for every minute of data scanned. The heart is calculated by taking the difference in time between consecutive R-peaks.
Early Myocardial Infarction Detection 43
Figure 18: R-peak Detection and PQSTJK extraction of ECG wave at level 24 4.
Detection of S, Q, T, T-peak, J, K, P point
Tompkins method for ST-segment analyzer was followed to compute J, T-peak, and Tpoint (Tompkins, 2000). The algorithm for estimating the Q and S point was derived from Tompkins method of ST-segment analyzer. After detecting R-peak, knowing the QRS complex duration to be 60ms, points Q and S were estimated as the first inflection points to the left and the right of R-peak respectively. After estimating the S-point, J-point was estimated to be first inflection point after S-point to the right of R-peak. T-peak was estimated to between R-peak+ 400ms to J-point+80ms. T-point was later on estimated from T-peak to T-point duration of 35ms to the R-peak side. Similarly K-point was estimated to be the first inflection point after Q on the left side of the R-peak, and P-point was estimated to be the first inflection point after K-point on the P-peak side. The detection of point is depicted in figure 18 above.
Early Myocardial Infarction Detection 44 5. Isoelectric line and ST-segment Computation After the estimation of all the relevant points in the ECG waveform, the isoelectric line and the ST-segment were computed. For the isoelectric line the mean value for point P and point K was computed and for ST-segment the mean value for point J and point T was computed. “Single ST-deviation as an absolute amplitude change between ST point and PR point values greater than 0.1mV” (Milosavljević Nebojša, et al.2006). The computed values for both isoelectric and the ST –segment were then compared within a range of ±0.1mV range. The complete code for the ST-segment change program is available in Appendix A section of this report. The generalized flow chart for the above described algorithm is given figure 21. The next section describes the testing and verification done using the datasets and the results obtained are also presented. 6. Graphical User Interface (GUI) A graphical user interface was created using Matlab’s GUIDE to create an intuitive interface to check for any abnormalities in an ECG data that is being tested. The ECG dataset name was given as the File name. After typing the File name, the start button activates the GUI program. The GUI consists of three axes figures. The first one from the top is original, unfiltered noisy signal that is being fed to the program. The second is the filetered and base line wander corrected signal. The third shows how the detection and estimation of parameters is being done in the program. The figures are refreshed after the every minute of procesing of the data. The GUI processes 5 minutes of the ECG dataset that is being fed to it. An ECG data that passes the test will show a “No ST Detected” message (figure 19) and an ECG data that has an ST-segment changes predicting abnormality, will show a “Warning” message (figure 20).
Early Myocardial Infarction Detection 45
Figure 19: Shows an intuitive GUI result for an ECG data with no MI detected
Figure 20: Shows an intuitive GUI result for an ECG data with an MI detected.
Early Myocardial Infarction Detection 46
Start
Load the.mat Data set File and the Header File Read the Header File and get required variables (Gain, Frequency)
Check the ECG Signal for 1minute
Filter the Data set and correct the baseline wanders
Detect PQRST Wave
Compute Isoelectric line (ISO) and ST segment
Is ST>ISO
YES
NO Go to the next minute’s signal
End
Figure 21: Flowchart for ST-elevation detection
Show MI Warning
Early Myocardial Infarction Detection 47
IX. Testing and Verification Extensive testing of the ST-change detection program was done to ensure the code estimates the ST-segment changes effectively. Receiver Operating Characteristic curve better known as ROC curve are “useful for organizing classifiers and visualizing their performance” (Fawcett Tom, 2006). ROC graphs are used for showing the tradeoff between the true positive rate and false positive rate. The next section discusses the results that were obtained during the testing. The total processing time for a minute of dataset was found to be ranging between 4 seconds to 10 seconds depending on the sampling frequency of the datasets. Calculation of Heart Rate every minute of data processing was done to account for the change in the heart rate. The testing and verification results are provided in the next section. a. Results A total of 54 datasets were tested. Random datasets from the five described databases were selected for the testing. The datasets were tested with the ST-change program and cross checked visually. The time interval for every dataset evaluated was one minute with a total of 5 minutes for each dataset. Out of every database close to 10 samples were analyzed. Out of the 54, 41 were True Positive whereas 13 were False Positive. 15 datasets from databases that have ST-change gave 100% accuracy and 11 datasets from normal sinus rhythm and noise stress showed no change in ST-segment i.e. again gave 100% accuracy rate. Other 15 datasets comprising of both the ST-change and the normal ECG database were accurate for 80%, 70%, 50% and 40% respectively. The other two accuracy levels 50% and 40% were chosen only for doing the ROC analysis. The overall accuracy of the algorithm was found to be approximately
Early Myocardial Infarction Detection 48 73% from the ROC analysis. Table 3 shows the details for the testing of databases. The detailed testing results are in Appendix B of this report. Table 3: Testing Results for ST-change detection program
Diagnostic Levels 90% 80% 70% 50% 40% Total
Observed Frequencies Cumulative Rates False True False True Postive Positive Positive Positive 4 26 0.3077 0.6341 1 4 0.3846 0.7317 2 4 0.5385 0.8293 1 4 0.6154 0.9248 5 3 1 1 13 41
A more detailed version of the database testing is available in Appendix C part of this report. The cumulative rates were calculated using the mathematical formulae developed by Lowry, Richard 2008. Table 4 shows the plotted point for ROC curve followed by the ROC curve in Figure 22. b. Discussion Based on the results it can be said with confidence that the ST- change detection program produces overall better results that many available method for ST measurement. Unlike many other ST –detection algorithms, this ST-change detection program also takes into consideration the change in the Heart Rate, which if ignored might not give precise ECG parameters estimation and result wrong ST-change predictions. Using Biorthogonal wavelets gives precise transformation points due to the similarity of shape between the ECG signal and that of the biorthogonal wavelets.
Early Myocardial Infarction Detection 49 Table 4: Data Points for ROC Curve False Positive Rate True Positive Rates (1-Specificity) Specificity) (Sensitivity) 0.05 0.1612 0.1 0.3553 0.15 0.4688 0.2 0.5494 0.25 0.6118 0.3 0.6629 0.35 0.706 0.4 0.7434 0.45 0.7764 0.5 0.8059 0.55 0.8326 0.6 0.857 0.65 0.8794 0.7 0.9001 0.75 0.9194 0.8 0.9375 0.85 0.9545 0.9 0.9705 0.95 0.9856
Figure 22:: Receiver Operating Characteristic Curve
Early Myocardial Infarction Detection 50 Use of Wavelet transforms speeds the signal processing of the ECG, which decreases the overall processing time for ST-segment estimation. Use of single level ECG parameter estimation reduces the time required for reconstruction of the signal. Also, faster dataset processing gives faster ECG parameter estimation, which gives faster response on the change in ST-segment, since the processing time for the ST-change detection program was found to be approximately 8 seconds. This shows that the fast algorithm of the ST-change detection program in real time ECG signal analysis will provide real time response, which is crucial in the case of myocardial infarction detection.
X. Economic Justification a. Executive Summary SMART Medical Devices Company is an established medical device company that design and sells medical diagnostic devices. It currently does not have a portable ECG device in the market. SMART Medical Device Company wants to develop a portable ECG device that can analyze physiological signals to detect the onset of myocardial infarction. Myocardial infarction is defined by the American Heart Association as the damaging or death of the heart muscles due to the blockage of blood supply. Diagnosing patients and detecting the symptoms, before the onset of myocardial infarction, is important to increase the mortality rate of patients. There are several ways to diagnose myocardial infarction and one of which is by using electrocardiogram (ECG) devices. Most of the ECG devices in the market are bulky and are not suitable for everyday use. In addition, most ECG devices are installed in hospitals where doctors or nurse can readily give diagnosis to patients. But, time is of the essence for doctors and nurses and they cannot be with the patients all the time. Furthermore,
Early Myocardial Infarction Detection 51 hospitals are loaded with patients that only need short-term care. In 2008 alone, there were more than 33 million short-term, acute patients that stayed in non-federal hospitals. (American Hospital Directory, 2008) One of the ways to reduce the numbers of short-term, acute patients staying in hospitals is by using a portable diagnostic device such as a small ECG device. It can be worn 24/7 with ease and comfort to patients. Healthcare providers can prescribe this device and unload hospitals with extra cost related to myocardial infarction disease. But to be useful, the portable ECG device must be smart and be able to process input data from the patients without the aid of a physician. The smart analytical tool that SMART Medical Devices develops complements the portable ECG device of the company since it can detect myocardial infarction without the aide of a doctor. Portable ECG devices are already in the market. Some are manufactured and marketed by big company competitors such as Philips Healthcare, Welch Allyn, and GE Medical. They gather ECG data from the patients and collect them to be sent by the patient to the doctor for further analysis. Some of them also have simple analytical tool to diagnose the patient. On the other hand, SMART Medical Devices’ ECG device has an algorithm that uses wavelet transform methodology as a smart analytical tool to analyze the ECG data coming form the patient. It can perform real-time analysis of the ECG data and can instantly notify the patient if there is an onset of myocardial infarction. Then a doctor can confirm the diagnosis remotely or as soon as the patient arrives in the hospital. In 2009, myocardial infarction had a prevalence of 8 million Americans and an estimated incidence of 900,000. Furthermore, coronary heart disease (CHD), which includes myocardial
Early Myocardial Infarction Detection 52 infarction, has a death rate of 144.4 in the United States. (American Heart Association, 2009) A smart portable ECG device with smart analytical tool can greatly reduce that numbers. For this project, a funding of $ 1.5 million is needed to start for the coding of the smart analytical software. The funding will be needed to acquire computer hardware and software licenses. Majority of the funds will go to the salary of the team, which composes of five software engineers, with different levels of expertise. It is estimated that the breakeven point will be reached on the third quarter of the product release, assuming the company will sell 500 units in the first year and with a price point of $2,000. b. Problem Statement The ability to detect the symptoms of myocardial infarction before the disease becomes severe is important to save the life of a patient. The best way to diagnose and identify the symptoms of myocardial infarction is ECG. Current ECG diagnostic devices do not have a smart analytical function to analyze ECG data. There’s a need to analyze ECG data from patients prone to myocardial infarction without the immediate aid of doctors or nurses. This can greatly increase the mortality rate of a patient and reduce the burden caused by myocardial infarction to hospitals. c. Solution and Value Proposition This project’s main focus is to develop an analytical tool that can analyze the ECG data of patients. The software can be installed in a portable ECG device developed by SMART Medical Devices. This project can greatly leverage the company to be the best in the industry. A well-made, and cost efficient portable ECG device that can accurately and efficiently diagnose patients can facilitate the realization of this mission.
Early Myocardial Infarction Detection 53 d. Market Size There’s a huge market for portable medical devices that diagnose heart diseas diseases. Heart disease is the number one cause of death in the United States. Every year, the American Heart Association estimates the prevalence of myocardial infarction. From 2003 to 2009, there was an estimated increase of 3.95% in the prevalence of myoc myocardial infarction in the United States (Seen ( in figure 23).
Figure 23: An estimate of myocardial infarction prevalence in the United States (Heart Disease and Stroke Statistics - 2003 Update, Heart Disease and Stroke Statistics - 204 Update, Heart Disease and Stroke Statistics - 2005 Update, Heart Disease and Stroke Statistics - 2006 Update, Heart Disease and Stroke Statistics - 2007 Update, Heart Disease and Stroke Statistics - 2008 Update, Heart Disease and Stroke Statistics - 2009 Update) And from the same period, there was an increase of 8.09% of new and recurrent attacks of myocardial infarction in the United St States (Seen in figure 24). The 2009 estimated direct and indirect cost of myocardial infarction was estimated to be $165 billion compared to $133 billion in 2004, a 24% increase (Seen in figure 25).
Early Myocardial Infarction Detection 54
Figure 24: An estimate of new and recurrent incidence of myocardial infarction in the United States(Heart Heart Disease and Stroke Statistics - 204 Update, Heart Disease and Stroke Statistics 2005 Update, Heart Disease and Stroke Statistics - 2006 Update, Heart Disease and Stroke Statistics - 2007 Update, Heart Disease and Stroke Statistics - 2008 Update, Heart Disease and Stroke Statistics - 2009 Update)
Heart Disease and Figure 25: The direct and indirect cost of myocardial infarction per year (Heart Stroke Statistics - 20044 Update, Heart Disease and Stroke Statistics - 2005 Update, Heart Disease and Stroke Statistics - 2006 Update, Heart Disease and Stroke Statistics - 2007 Update, Heart Disease and Stroke Statistics - 2008 Update, Heart Disease and Stroke Statistics - 2009 Update)
Early Myocardial Infarction Detection 55 In 2009, there are a total of 7.9 million cases of myocardial infarction in the United States. Every 34 seconds, an American will suffer a heart attack, and from that, 25% men and 38% women will die within the first year after suffering a heart attack. In 2005, the recorded number of deaths due to myocardial infarction was more 150,000 Americans (American Heart Association, 2009). Giving patients, who are prone to heart attacks or who had suffered an initial heart attack, an access to a portable ECG device can greatly decrease their mortality rate. The US market for home monitoring is expected to increase over 5% annually to $1.8 Billion in 2012 (Demand for Home Medical, n.d.). SMART Medical Devices will market the MI Detector, a portable ECG device with the smart algorithm, in the United States. e. Competitors Heart monitoring device is a big market, considering that these devices diagnose the number one cause of death in the United States. With such, there are many competitors in this industry. Philips Healthcare, a part of Koninklijke Philips Electronics N.V. (Royal Philips Electronics), developed an ECG Holter device, DigiTrak XT. It can monitor and record the patient’s ECG data for up to 7 days. The recorded data are then transferred to personal computer using Philips Healthcare’s software and analyzed before sending the data securely sent to a data center. The Philips Healthcare DigiTrak XT Holter device can cost almost $9,000. Welch Allyn, another company that develops ECG Holter devices, has HR 100 and HR 300 ECG Holter devices that record, and store patient’s data. The recorded data are also downloaded to a personal computer and analyzed using proprietary algorithm to detect any abnormalities in the patient’s recorded ECG data. It is sold online and cost approximately $3,000.
Early Myocardial Infarction Detection 56 GE Healthcare, a part of General Electric Company, developed SEER Light compact digital recorder, a compact ECG Holter device that can record ECG data for up to 48 hours. It is a part of GE Healthcare’s ambulatory system to analyze the patient’s ECG data for any anomalies. After monitoring the patient, the data is downloaded and analyzed using an advance algorithm and can be stored to a central database, which is run by proprietary software. The device can be purchased for almost $9,000. Omron Healthcare has a portable ECG Monitor, HCG-801, which can be used by patients whenever symptoms of heart disease occur. It has a large screen that can display the ECG data from the patient, but the data also needs to be downloaded to a computer for analysis. The device is cheap and is found online for almost $500. All of the devices mentioned are ECG Holter devices that can monitor the patient’s heart activity for 24 hours or more. They are portable and can be worn for a long period of time without impeding day-to-day activities of patients. The portable and easy to use MI Detector device by SMART Medical Devices will be a Holter device, which can provide constant and real time analysis of ECG devices. The advantage of the MI Detector over the competitors’ is that it can monitor the patient and process the data recorded in the portable ECG device without downloading the data to a personal computer. f. Customers The target markets for SMART Medical Device’s MI Detector are physicians that will prescribe the ECG monitoring device to patients. This ECG device is considered a “prescription” device since it monitors physiological functions of the patient. Special training needs to be administered to the patient before a patient can bring the device home. The device will be particularly marketed to doctors specializing in heart conditions. And since this device
Early Myocardial Infarction Detection 57 will be regulated under the Food and Drug Administration (FDA) law, this device will be marketed initially in the United States. The device can be purchased by individual patients in pharmacies or through online channel, granted they have a prescription from their physician. The MI Detector device can be covered by Medicare or Health Insurance or can be purchased, out-of-pocket, at a manufacturer suggested retail price (MSRP). g. Cost There are several factors that influence the cost of MI Detector. The MI Detector device is composed of a hardware and software components. The hardware component is outsourced to an Original Equipment Manufacturer (OEM), which has a proven record for manufacturing quality-made medical devices. The software component is developed in-house and is designed by SMART Medical Device’s software department. The total cost in designing, developing, marketing, delivering and servicing the MI detector is estimated from different cost drivers. i. Fixed costs Fixed costs are expenses that do not change and are not based on the activity to develop and market the device. Majority of the fixed cost associated with the MI Detector consists of salary of software engineers. The software development team is composed of two level 1 or 2 software engineers, three level 4 to 5 software engineers, and a department manager. Other costs associated with the fixed costs of MI Detector device are: the purchase of software licenses and hardware components as well as other development cost for improving the algorithm of the MI Detector. The fixed cost of MI Detector is listed in table 5.
Early Myocardial Infarction Detection 58 ii.
Variable Costs
Variable costs are cost drivers that are associated with the activity from developing to servicing the MI Detector device. It varies from time to time due to the number of volume of producing and selling the device. Since an OEM vendor manufactures the MI Detector device, there is no associated cost with manufacturing. Logistics and manufacturing contracts are major cost drivers for this device, which consists of contracting the manufacturing to an OEM manufacturer, receiving and delivery of the device from the OEM manufacturer to the sales channel, and keeping the inventory in order. Other cost drivers are sales and marketing, the support staff and the initial regulatory and legal requirements to manufacture and sell the device. Table 6 illustrates the variable costs of MI Detector device. Table 5: Fixed cost of MI Detector device Fixed Cost
2009
2010
2011
2012
2013
$6,000.00
$1,000.00
$1,000.00
$1,000.00
$1,000.00
Level 1/2 Engineers (2)
$100,000.00
$105,000.00
$110,250.00
$115,762.50
$121,550.63
Level 4/5 Engineers (3)
$300,000.00
$315,000.00
$330,750.00
$347,287.50
$364,651.88
Department Manager
$120,000.00
$126,000.00
$132,300.00
$138,915.00
$145,860.75
Total S. Engr Salary
$520,000.00
$546,000.00
$573,300.00
$601,965.00
$632,063.26
Development Cost
$50,000.00
$0.00
$0.00
$0.00
$0.00
Business Expense
$100,000.00
$100,000.00
$100,000.00
$100,000.00
$100,000.00
Total Fixed Cost $676,000.00
$647,000.00
$674,300.00
$702,965.00
$733,063.26
Hardware and License Software Engineer Salary
Early Myocardial Infarction Detection 59 h. Price Point The price of the MI Detector device will be below the competitor’s prices. The MI Device will be sold with an MSRP of $2,000, which is well below the competitor’s. The device, when purchased initially, is a complete device since all the analytical processing is done in the device. Occasional accessory replacement is necessary after every use to ensure accurate ECG data recording and analysis.
Table 6: Variable cost of MI detector device Variable Cost
2009
2010
2011
Marketing Salary
$150,000.00
$157,500.00
$165,375.00
$173,643.75
$182,325.94
Support Staff
$150,000.00
$157,500.00
$165,375.00
$173,643.75
$182,325.94
Logistics and Contracts
$500,000.00
$500,000.00
$500,000.00
$500,000.00
$500,000.00
Regulatory and Legal
$50,000.00
$5,000.00
$5,000.00
$5,000.00
$5,000.00
$820,000.00
$835,750.00
$852,287.50
$869,651.88
Total Variable Cost $850,000.00
2012
2013
i. SWOT Assessment The device itself has much strength and some weaknesses. The SWOT assessment of the MI Detector device is shown in table 7.
Early Myocardial Infarction Detection 60 Table 7: SWOT assessment of MI Detector medical device Strengths
• Provides real-time and accurate
Weaknesses •
analysis of ECG data • Data does not need to be downloaded for processing • Portable and easy to use • Can be worn several hours per day Opportunities •
There are currently no other • major medical device company that offers the same product that we offer
Since the device is a product in the company portfolio, patients may be hesitant to try and use the product
Threats Other medical device companies might develop similar products with similar price point
j. Investment Capital Requirement The MI Detector device is composed of two components: Hardware and Software. Since the MI Detector device’s major component is the smart software installed to diagnose myocardial infarction, majority of the funding will be focused on the software development team. In addition, SMART Medical Devices will outsource the manufacturing of the portable ECG device to an OEM company. Thus, the capital required to fund the project will primarily be divided down for software engineers’ salaries, and the logistics of the portable device. The initial year will require a budget of $1.5 million to fund the initial development of the product. The company is expected to become profitable with this device on the second year of developing and selling the device, as shown in figure 26. The company is expected to become profitable with this device on the second year of developing and selling the device, as shown in figure 27.
Early Myocardial Infarction Detection 61
Initial Investment $1,800,000 $1,600,000 $1,400,000 Costs
$1,200,000 $1,000,000 $800,000 $600,000 $400,000 $200,000 $0 Fixed Cost
Variable Cost
Investment
Figure 26: Initial Investment Requirement of MI Detector device
Yearly Model Revenue
Total Cost
Profit and Loss
$5,000,000 $4,000,000 Sales
$3,000,000 $2,000,000 $1,000,000 $0 -$1,000,000
1
2
3
4
5
Year
Figure 27 27: Yearly Model of MI Detector device Having a price point of $2,000 and an estimate of 500 units sold on the first year, the company will have a loss of $500,000 on the first year. With an estimated sales increase of 20% every year or 5% every quarter, the company is estimated to gain profit of almost $2.5 million on the fifth year. (Shown in table 8)
Early Myocardial Infarction Detection 62 Table 8: The break-even table of MI Detector device 2009 2010 2011 2012 2013 520 1200 1440 1730 2100 Number of Units sold $2,000 $2,000 $2,000 $2,000 $2,000 Price $1,040,000 $2,400,000 $2,880,000 $3,460,000 $4,200,000 Revenue $676,000 $647,000 $674,300 $702,965 $733,063 Fixed Cost $850,000 $820,000 $835,750 $852,288 $869,652 Variable Cost $1,526,000 $1,467,000 $1,510,050 $1,555,253 $1,602,715 Total Cost -$486,000 $933,000.00 $1,369,950.00 $1,904,748 $2,597,285 Profit and Loss
k. Personnel This project only justifies the requirements of the software development team of SMART Medical Devices. This project is required to develop the smart analytical software of the MI Detector device. The software development team of the company consists of five software engineers with varying levels, and a department manager. The members of the development team are currently employed at SMART Medical Devices. A department manager will oversee the development of the project and needs to have skills and knowledge to drive the project forward. The department manager should have an experience in project management and can deal well with different levels of the company. There are three level 4 to 5 software engineers with the following minimum qualifications to develop the project:
• At least 5 years experience • Have a BS or MS degree in Computer Engineering/Electrical Engineering/Mechanical Engineering/Biomedical Engineering
• Experience in hardware and software development
Early Myocardial Infarction Detection 63 Two level 1 or 2 software engineers are also employed for this project to help in developing the algorithm of the MI Detector. Minimum qualifications for these positions are the following:
• 1 or 2 years experience in software and/or hardware projects • BS Degree in Computer Engineering/Electrical Engineering/Mechanical Engineering/Biomedical Engineering
• Experience in software language • Able to work without supervision l. Business and Revenue Model SMART Medical Devices have an established sales and marketing team. The device will be sold to patients through direct and indirect sales. The device will be marketed initially in the United States after obtaining the necessary permit from the FDA. SMART Medical Devices will use direct marketing techniques through advertising and attending conventions and events. SMART Medical Devices will also utilize the well-established sales force of the company and market the device through contract sales. The company website will also feature the MI Detector device and consumers will be able to purchase the device online. Marketing materials will also be posted to educate consumers on the features and benefits of MI Detector. The company will approach insurance companies, for greater acceptance by the consumer and having an insurance reimbursement codes will greatly increase the chances of selling the device. Even though the device will be covered by the insurance, consumers will be willing to pay, out-of-pocket, for the MI Detector due to the great benefits it can provide.
Early Myocardial Infarction Detection 64 The company elected to manufacture the device through an OEM company that are reputed for their quality-made medical devices. Having an OEM company manufacturer, the device is expected to minimize the cost and increases profit to the company. The MI Detector will have an MSRP of $2,000 which is comparable or well below the price of the competitor. Other accessories of the device will also be sold with a less or comparable prices. m. Strategic Alliances/Partners SMART Medical Device Company will partner with physicians to develop the MI Detector. Having the inputs and comments of doctors during the design phase can greatly leverage the device to be the best diagnostic device in diagnosing myocardial infarction. Partnership with a well-known hospital can be a great promotional tool and can provide good benefits in both ends. Consumers take medical advices from well known institutions and medical institutions can give a technological achievement when a medical device proves to be effective and efficient. Also, a well-established partnership with the OEM manufacturer is in place to deliver the best diagnostic device in the market. n. Profit and Loss The company will outsource the manufacturing of the MI Detector ECG device to minimize its expenses. Majority of the fixed costs are attributed to the salary of the software development team. It will also cover for the expenses to develop the smart algorithm of the ECG device. Those expenses consist of the hardware equipments needed and other business expenses such as utilities and rent. The rest of the expenses are expected to be variable, giving the company to expand as needed.
Early Myocardial Infarction Detection 65 i. Demand Assumptions The SMART Medical Devices considers 100% of all patients with myocardial infarction as a potential customer of MI Detector device.
• In 2009, there is an estimated 7.9 million Americans with myocardial infarction condition
• An annual incidence of 610,000 and 325,000 new and recurrent attacks of myocardial infarction
• Every 34 seconds, an American will suffer myocardial infarction • 25% of men and 38% of women die within the first year of having the diagnosed with myocardial infarction; 50% of men and women age under 65, will die within 8 years. ii. Product Assumptions In the initial quarter, no products will be sold in the market, since the MI Detector device is still in development phase. It is expected to be in the market on the third quarter of 2009. An initial sale of 500 units in the first year is expected and is conservatively anticipated to grow 5% per quarter or 20% per year. Table 9 and figure 28, illustrates the quarterly model with 5% quarterly growth. With the assumption that there are no unforeseen circumstances, the company is expected to break even on the third quarter of its release. On the second year, it is expected that the demand will pick up and will see the continued growth and success of MI Detector ECG device.
Early Myocardial Infarction Detection 66 Table 9: The quarterly model of SMART Medical Devices Q2 2009 Number of Units sold Price Revenue Fixed Cost Variable Cost Total Cost Profit & Loss
0
Q3 2009 250
Q4 2009 265
Q1 2010 275
Q2 2010 290
Q3 2010 305
Q4 2010 320
$2,000.00 $0.00 $225,000.00 $285,000.00
$2,000.00 $500,000.00 $225,000.00 $285,000.00
$2,000.00 $530,000.00 $225,000.00 $285,000.00
$2,000.00 $550,000.00 $225,000.00 $205,000.00
$2,000.00 $580,000.00 $225,000.00 $205,000.00
$2,000.00 $609,000.00 $225,000.00 $205,000.00
$2,000.00 $639,450.00 $225,000.00 $205,000.00
$510,000.00 -$510,000.00
$510,000.00 -$10,000.00 $10,000.00
$510,000.00 $20,000.00
$430,000.00 $120,000.00
$430,000.00 $150,000.00
$430,000.00 $179,000.00
$430,000.00 $209,450.00
Break Even Analysis Revenue
Total Cost
Profit & Loss
$800.00
$ Thousands
$600.00 $400.00 $200.00 $0.00 -$200.00
1
2
3
4
5
6
7
-$400.00 -$600.00
Quarterly
Figure 28: The quarterly model of SMART Medical Devices o. Exit Strategy The company will lose revenue on the first year due to the development phase and initial introduction of MI Detector to the market. The company is estimated to be profitable from the first year to the fifth year, and beyond. Even though the company expects to be profitable on the third quarter of selling the device, unforeseen circumstances may arise and can alter the financial forecast of the company. If the company ny does not make any profits from m this device, SMART Medical Devices’ management may elect to discontinue marketing and selling the device.
Early Myocardial Infarction Detection 67 Since SMART Medical Devices developed a smart analytical tool to analyze ECG data, the company will sell the copyrights of the software algorithm to established portable ECG manufacturing company to get back the investment made for MI Detector. The company may also elect to market only the smart analytical software to patients with existing portable ECG devices. Having smart analytical software installed in their portable ECG devices can greatly increase the efficacy of those machines to detect myocardial infarction. The other appropriate exit stratergy will be market the smart algorithm as add-on for computer programs and mobiles and market and sell the smart algorithm as a healthcare application.
XI. Future Directions We have developed a ST-change detection program that can accurately analyze ECG data and can determine if it contains any traits of ST-segment elevation or depression and based on the analysis predicatively warn onset of myocardial infarction. This project, we believe, can greatly decrease the morbidity and mortality of patients that are prone to have myocardial infarction, as well as, those that are suffering from it. Although the project is successful in diagnosing myocardial infarction, there are still many issues that need to be addressed. The project was developed using datasets gathered from the Physiobank database. Applying the project directly to patients and the ability to analyze data in real-time can help us better understand the myocardial infarction condition which can lead to improved analytical decisions. This can be done using the Real Time Toolbox that is present in Matlab 7.1 software, the connection to the ECG machine can be made through a USB port and the real-time data can be captured and analyzed. With an addition of few hardware components it will be possible to have a standalone bedside ST-change monitoring device.
Early Myocardial Infarction Detection 68 The project was designed using the Matlab software. The group wanted to port the program to a portable ECG device that can be comfortably worn by patients for a prolonged period of time. This can effectively and efficiently diagnose the patient and eliminate missed diagnosis. Another future consideration is by having wireless communication capability to the portable ECG device. Personal area network (PAN) such as Bluetooth was considered to create a wireless probe that connects wirelessly to a base ECG device. The ECG device is then connected wirelessly to the hospital or doctor through wide area network (WAN) such wireless internet or WI-FI. Telecardiology is rising in popularity and incorporating wireless internet to the ECG device can provide long distance diagnosis to the patients by their physician and give patients the freedom to roam outside the reach of hospitals. With telecardiology in mind, an ECG with a GPS feature can give patients an additional security.
XII. Conclusion The algorithm created in MATLAB was successful in detecting the different segments of ECG signal from the Physiobank database. The QRS complex was detected and was used to identify the ST-segment. The code was able to detect the abnormalities in the ST segment with high accuracy. It was also successful in eliminating noises and baseline drifts that can degrade the accuracy of the algorithm. With the use of biorthogonal wavelets the ECG signal processing was made faster so that when real time ECG signal is fed to the algorithm, the processing of the ECG signal and the resultant warning in the case of abnormality will be close to the actual signal. The algorithm was successful in identifying ST-segment changes/abnormalities for single lead ECG signal.
Early Myocardial Infarction Detection 69 Incorporating this algorithm into a 12-lead ECG monitoring system will make a standalone myocardial infarction detection device.
Early Myocardial Infarction Detection 70 XIII. References Albrecht P. (1983). “S-T segment characterization for long-term automated ECG analysis”. M.S. Thesis, MIT Dept. of Electrical Engineering and Computer Science, (1983) American Heart Association. (2002). Heart Disease and Stroke Statistics - 2003 Update. Dallas, TX: American Heart Association. American Heart Association. (2003). Heart Disease and Stroke Statistics - 2004 Update. Dallas, TX: American Heart Association. American Heart Association. (2005). Heart Disease and Stroke Statistics - 2005 Update. Dallas, TX: American Heart Association. American Heart Association. (2006). Heart Disease and Stroke Statistics - 2006 Update. Dallas, TX: American Heart Association. American Heart Association. (2007). Heart Disease and Stroke Statistics - 2007 Update. Dallas, TX: American Heart Association. American Heart Association. (2008). Heart Disease and Stroke Statistics - 2008 Update. Dallas, TX: American Heart Association. American Heart Association (2008). “What is a Heart Attack?” Life after Heart Attack, Diseases and Conditions, American Heart Association (2008). Retrieved on November 6th 2008 from http://www.americanheart.org/presenter.jhtml?identifier=3038238 American Heart Association. (2009). Heart Disease and Stroke Statistics - 2009 Update. Dallas, TX: American Heart Association. American Hospital Directory - Hospital Statistics by State. (2008, August 6). Retrieved March 28, 2009, from http://ahd.com/state_statistics.html Barron Jon (2007). “Secrets of the Heart”. Baseline of Health Foundation (2007). Retrieved on 6th April 2009 from http://www.jonbarron.org/heart-health-program/07-02-2007.php Bhatia Praval, Boudy Jerome, Varejao Rodrigo (2006). “Wavelet transformation and preselection of mother wavelets for ECG signal processing” Proceedings of the 24th IASTED International Multi-conference. Biomedical Engineering February 2006. Cardiovascular Consultants (2006). “ Physiology”. Retrieved on March 28th 2009 from http://www.cardioconsult.com/Physiology/ Coronary Artery Disease (2008). Up to Date. Retrieved on 07th December 2008 from http://www.uptodate.com/patients/content/images/card_pix/Coronary_artery_disease.jpg
Early Myocardial Infarction Detection 71 Demand for Home Medical Equipment to Exceed $10 b. (n.d.). Retrieved March 28, 2009, from www.expresshealthcaremgmt.com/200903/market16.shtml Fawcett Tom (2005). “An introduction to ROC Analysis”. Pattern Recognition letters 2006, Issue # 27. Pages: 861-874. Retrieved on April 3rd 2009 from http://www.csee.usf.edu/~candamo/site/papers/ROCintro.pdf Franc Jager, Alessandro Taddei, George B. Moody, Michele Emdin, Gorazd Antolic, Roman Dorn, Ales Smrdel, Carlo Marchesi, and Roger G. Mark. Long-term ST database: a reference for the development and evaluation of automated ischaemia detectors and for the study of the dynamics of myocardial ischaemia. Medical & Biological Engineering & Computing 41(2):172-183 (2003) Goldberger, A. L.,Amaral, L. A. N.,Glass, L., Hausdorff, J. M. and Ivanov, P. Ch.,Mark, R. G., Mietus, J. E., Moody, G. B., Peng, C.-K., Stanley, H. E.(2000). “{PhysioBank, PhysioToolkit, and PhysioNet}: Components of a New Research Resource for Complex Physiologic Signals" June 13th 2000. Journal: Circulation, Volume: 101, Issue # 23. Pages: e215--e220. Retrieved on 22nd November 2008 from http://circ.ahajournals.org/cgi/content/full/101/23/e215 Heart Information Center (2006). “Anatomy of Heart”. Texas Heart Institute (July 2006). Retrieved on 17th March 2009 from http://www.texasheart.org/HIC/ANATOMY/anatomy2.cfm Heart Information Center (2006). “Bundle Branch Block”. Texas Heart Institute (February 2009). Retrieved on 25th March 2009 from http://www.texasheart.org/HIC/Topics/Cond/bbblock.cfm Jouck. P.P.H. (2004). “Application of the Wavelet Transform Modulus Maxima to the T-wave Detection in Cardiac Signals” December 2004. Retreived on 17th March 2009 from http://www.personeel.unimaas.nl/Westra/PhDMaBateaching/GraduationStudents/PJouck2004/PJouck2004verslag.pdf Klabunde, R. E. (2007, April 6). ECG Introduction. Retrieved April 4, 2009, from http://www.cvphysiology.com/Arrhythmias/A009.htm Klatt E.C. (2008). “Myocardial Infarction” The University of Utah Eccles Health Sciences Library (2008). Retrieved on 29th March 2009 from http://library.med.utah.edu/WebPath/TUTORIAL/MYOCARD/MYOCARD.html Lowry, Richard (2008). “Simple ROC Curve Analysis: Version 1”. VassarStat: Web Site for Statistical Computation 2001-2009. Retrieved on April 3rd 2009 from http://faculty.vassar.edu/lowry/roc1.html#down Milosavljević Nebojša., Petrović Aleksandar. (2006). “ST Segment Change Detection by Means of Wavelets”. 8th Seminar on Neural Networks Applications in Electrical Engineering, NEUREL-2006. http://www.ewh.ieee.org/reg/8/conferences.html.
Early Myocardial Infarction Detection 72
Moody GB, Muldrow WE, Mark RG. A noise stress test for arrhythmia detectors. Computers in Cardiology 1984; 11:381-384. Myocardial Infarction (2009). “Myocardial Infarction”. The Free Dictionary by Farlex. Retrieved on 29th March 2009 from http://medicaldictionary.thefreedictionary.com/Myocaridal+infarction">myocardial infarction O'Grady, M. R., & O'Sullivan, M. L. (n.d.). VetGo Cardiology. Retrieved April 4, 2009, from http://www.vetgo.com/cardio/concepts/concsect.php?conceptkey=165 Pan Jiapu., Tompkins Willis. J. (1985). “A Real-Time QRS Detection Algorithm”. IEEE Transactions on Biomedical Engineering, Vol. BME-32, Issue #3, March 1985, Pages:230-236. Polikar Robi (2001). “The Engineers ultimate guide to wavelet analysis”. The Wavelet Tutorial. College of Engineering, Rowan University (2001). Retrieved on 5th April 2009 from http://users.rowan.edu/~polikar/WAVELETS/WTtutorial.html Poul-Erik Paulev(2000). “Textbook in Medical Physiology and Pathophysiology Essentials and Clinical Problems- Chapter 10: Cardiac Performances and Disorders”. Copenhagen Medical Publishers 1999-2000. Retrieved on 29th March 2009 from http://www.mfi.ku.dk/ppaulev/chapter10/chap_10.htm Samer Garas, Maziar Zafari (2008). “Myocardial Infarction”. EMedicine from WebMD (August 2008). Retrieved on 27th March 2009 from http://emedicine.medscape.com/article/155919overview Saritha.C., Sukanya. V., Murthy. Narasimha. Y. (2008) . “ECG Signal Analysis using Wavelet Transforms”. Journal of Bulg. J. Physics(2008). Issue 35. Pages: 68-77. Schniter Phil (2005). “Continous Wavelet Transform”. Creative Commons Attribution License. Connexious module: m10418. Version: 2:13. June 9th, 2005. Retrieved on April 5th 2009 from http://cnx.org/content/m10418/latest/ Sivannarayana N., Reddy D.C. (1999). “Biorthogonal Wavelet Transforms for ECG Parameters Estimation”. Journal Medical Engineering and Physics (1999) Issue: 21Pages: 167-174. Taddei, A., Distante, G., Emdin, M., Pisani, P., Moody, G.B., Zeelenberg, C., and Marchesi, C. The European ST-T Database: standard for evaluating systems for the analysis of ST-T changes in ambulatory electrocardiography. European Heart Journal 13: 1164-1172 (1992). The Myocardial Infarction –Heart Attack (2006). The Causes: The Myocardial Infarction: The Cardiovascular Diseases. Heart and Vessels (July 2006). Retrieved on 29th March 2009 from http://www.heart-vessels.com/cardiovascular-diseases/heart-attack-myocardialinfarction6.php
Early Myocardial Infarction Detection 73
Tompkins Willis. J. (2000). “Biomedical Digital Signal Processing”. Chapter 13: ECG Analysis System. Prentice Hall (2000). Pages: 271-272. VIBES Electrocardiogram Mosaic. (n.d.). Retrieved April 4, 2009, from http://www.vanth.org/vibes/electro.html Wenli Chen., Zhiwen Mo., Wen Guo. (2007) “Detection of QRS Complexes Using Wavelet Transforms and Golden Section Search” Advances in Intelligent Systems Research ISKE2007 Proceedings. Retrieved on March 18th 2009 from http://www.atlantispress.com/php/download_paper.php?id=1195. Wolfram Research (2009) . “Biorthogonal Wavelets”. Documentation: Wavelet Explorer: Fundamentals of Wavletes. Retrieved on April 8th 2009 from http://documents.wolfram.com/applications/wavelet/FundamentalsofWavelets/1.6.1.html World Heart Federation (2009). Cardiovascular Diseases. http://www.world-heart-federation.org/cardiovascular-health/heart-disease/different-heartdiseases/#c434 Yanowitz G.Frank (2006). “IX. Myocardial Infarction”. The Alan E. Lindsay ECG Learning Center in Cyberspace. Spencer S. Eccles Health Sciences Library University of Utah Health Services Center (2006). Retrieved on November 6th 2008 from http://library.med.utah.edu/kw/ecg/ecg_outline/Lesson9/index.html#Intro
Early Myocardial Infarction Detection 74 Appendix A
Code for Matlab Program clear all; clc; close all; %*************************************************************** % Get Data & User Inputs %*************************************************************** Fileloc = 'C:\MATLAB7\Work\'; Filename = input('Enter ECG File Name = ','s'); % Input Filename Headerfile = strcat(Filename,'.hea');
% Header In TXT Format
load (Filename);
% .mat file for Data
%*************************************************************** % Load Header Data %*************************************************************** fprintf(1,'\nK> Loading Data from Header File %s ...\n', Headerfile); signalh = fullfile(Fileloc, Headerfile); fid1 = fopen(signalh,'r'); z = fgetl(fid1); A = sscanf(z, '%*s %d %d',[1,2]); % Number Of Signals
nosig = A(1); sfreq = A(2); clear A; z = fgetl(fid1); A = sscanf(z, '%*s %*d %d %d %d %d',[1,4]); gain = A(1); clear A;
% Integers Per mV
Early Myocardial Infarction Detection 75 S = sfreq*60; counter1=0; counter2=0; counter3=0; for n = 0:5 tic j = S*n+1:1:S*(n+1); D = val(j); dat = length (D); k = 1:1:dat; D = D(k)/gain; %*************************************************************** %Signal filter and Base line wander correction %*************************************************************** D= transpose (D); windowSize = 5; filsig = filter (ones(1,windowSize)/windowSize,1,D); y = medfilt1(filsig,200); % 1st median filter s1 = y; clear y; y = medfilt1(s1,600); % 2nd median filter D = filsig - y; clear s1; clear y; pack;
%*************************************************************** % Manipulate Data So We Only Look At What The User Wants %*************************************************************** D = transpose (D);
Early Myocardial Infarction Detection 76 D = cwt (D, 1:4, 'bior2.4'); %Performing Continuous Wavelet Transform using Biorthogonal Wavelet to ECG_1 D = transpose (D); x = D (:,4); clear D; %*************************************************************** % R-Peak Detection %*************************************************************** thresh = 0.6; % create time axis len = length (x); tt = 1/sfreq:1/sfreq:ceil(len/sfreq); t = tt(1:len); max_h = max (x(round(len/4):round(3*len/4)));%segment search area first find the highest bumps in the ECG_1 poss_reg = x>(thresh*max_h); %then build an array of segments to look in %find indices into boudaries of each segment left
= find(diff([0 poss_reg'])==1); % remember to zero pad at start
right = find(diff([poss_reg' 0])==-1); % remember to zero pad at end %loop through all possibilities for(i=1:length(left)) [maxval(i) maxloc(i)] = max( x(left(i):right(i)) );
Early Myocardial Infarction Detection 77 [minval(i) minloc(i)] = min( x(left(i):right(i)) ); maxloc(i) = maxloc(i)-1+left(i); % add offset of present location minloc(i) = minloc(i)-1+left(i); % add offset of present location end R_index = maxloc; R_t
= t(maxloc);
R_amp = maxval; %*************************************************************** % Heart Rate Calculation %*************************************************************** for j = 2:length (R_t) HR(j)= R_t(j)-R_t(j-1); end H_R = 60/(mean (HR)); fprintf (1,'\nK> Heart Rate is %d \n',H_R); %*************************************************************** % S-Point Detection %*************************************************************** R_len= length (R_index); for j = 1:R_len IR1 = R_index(j); for i = IR1:IR1+ (round(sfreq*0.03)*(H_R/72)) if i == length (x)|i==0 S_index(j)= 1; S_amp(j) = x(1,1); S_t(j) = t(1,1); break end
Early Myocardial Infarction Detection 78 if x(i,1)< x(i+1,1) && x(i,1)< x(i-1,1) S_index(j)= i; S_amp(j) = x(i,1); S_t(j) = t(1,i); break end end end %*************************************************************** % Q-Point Detection %*************************************************************** for j = 1:R_len IR1 = R_index(j); for i = IR1:-1:IR1- (round(sfreq*0.03 *(H_R/72))) if i == 0|i==length (x) Q_index(j)= 1; Q_amp(j) = x(1,1); Q_t(j) = t(1,1); break end if x(i,1)< x(i+1,1) && x(i,1)< x(i-1,1) Q_index(j)= i; Q_amp(j) = x(i,1); Q_t(j) = t(1,i); break end end end
Early Myocardial Infarction Detection 79 %*************************************************************** % J-Point Detection %*************************************************************** S_len = length (S_index); for j = 1:S_len IS1= S_index(j); for i=IS1:IS1+ (round(sfreq*0.03) *(H_R/72)) if i==0 J_index(j)=1; J_amp(j)= x(1,1); J_t(j)= t(1,1); break end if i> length (x) break end if x(i,1)>=0 J_index(j)=i; J_amp(j)= x(i,1); J_t(j)= t(1,i); break end end end %*************************************************************** % T-Peak Detection %*************************************************************** J_len = length (J_index); for j = 1: J_len P1 = R_index(j)+ (round(sfreq*0.4) *(H_R/72));
Early Myocardial Infarction Detection 80 P2 = J_index (j)+ (round(sfreq*0.08) *(H_R/72)); if P1> length (x)|P2> length (x) break end if P1 > P2 [T_peak(j),T_peak_index(j)] = max(x(P2:P1)); T_peak_index(j) = T_peak_index(j)+ P2; else [T_peak(j),T_peak_index(j)] = max(x(P1:P2)); T_peak_index(j) = T_peak_index(j)+ P1; end T_peak_t (j)= t(T_peak_index (j)); end
%*************************************************************** % T-Point (T onset) Detection via T-peak %*************************************************************** T_len = length (T_peak_index); for j = 1:T_len IT1 = T_peak_index(j); for i = IT1:-1:IT1-(round(sfreq*0.035) *(H_R/72)) TP_index(j)=i; TP_amp(j)=x(i,1); TP_t(j)=t(1,i); end end
Early Myocardial Infarction Detection 81 %*************************************************************** % K-Point %*************************************************************** Q_len = length (Q_index); for j = 1:Q_len IQ1 = Q_index(j); for i=IQ1:-1:IQ1- (round(sfreq*0.03) *(H_R/72)) if i == 0 K_index(j) = 1; K_amp(j)= x(1,1); K_t(j)=t(1,1); break end if x(i,1)>=0 K_index(j) = i; K_amp(j)= x(i,1); K_t(j)=t(1,i); break end end if i == 0 K_index(j) = 1; K_amp(j)= x(1,1); K_t(j)=t(1,1); break end if x(i,1)< x(i+1,1) && x(i,1)< x(i-1,1) K_index(j) = i; if K_index(j)==0
Early Myocardial Infarction Detection 82 K_index(j)=1; end K_amp(j)= x(i,1); K_t(j)=t(1,i); end
end
%*************************************************************** % P-Point Detection via K+80ms %*************************************************************** K_len = length (K_index); for j = 1:K_len IK1 = K_index(j); for i=IK1:-1:IK1- (round(sfreq*0.08) *(H_R/72)) if i ==0 P_index(j) = 1; P_amp(j)= x(1,1); P_t(j)=t(1,1); break end P_index(j) = i; P_amp(j)= x(i,1); P_t(j)=t(1,i); end end
Early Myocardial Infarction Detection 83 %*************************************************************** % Calculation of Isoelectric Line %*************************************************************** j = 1:1:K_len; ISO(j) = mean(x(P_index(j):K_index(j))); %*************************************************************** % Calculation of ST-segment %*************************************************************** a = length (J_index); b = length (TP_index); if a==b; j = 1:1:J_len; ST(j) = mean(x(J_index(j):TP_index(j))); end if a>b j = 1:1:b; ST(j) = mean(x(J_index(j):TP_index(j))); end if a= (ST(j)+0.0001) && ISO(j)>=ST(j)0.0001|ISO(j)==ST(j)
Early Myocardial Infarction Detection 84 counter2=counter2+1; else counter3=counter3+1; end end end
if a=ST(j)+0.0001 && ISO(j)>=ST(j)0.0001|ISO(j)==ST(j) counter2=counter2+1; else counter3=counter3+1; end end end
if a>b for j=1:b counter1=counter1+1; if ISO(j)>=ST(j)+0.0001 && ISO(j)>=ST(j)0.0001|ISO(j)==ST(j) counter2=counter2+1; else counter3=counter3+1; end
Early Myocardial Infarction Detection 85 end end clear ISO; clear ST; toc fprintf(1,'\nK> %d loop completed %n \n',n); end fprintf (1,'\nK> total number of signals evaluated is %d \n',counter1) fprintf (1,'\nK> total number of signals without MI is %d \n',counter2) fprintf (1,'\nK> total number of signals with MI is %d \n',counter3) if counter3/counter1>=0.95 fprintf(1,'\nK>WARNING: MI\n'); else fprintf(1,'\nK>No MI\n'); end %*************************************************************** %Plotting Function %*************************************************************** figure subplot(2,1,1) plot(t,x), grid on; title('Level 4 2^4 Biorthogonal Wavelet Transformed ECG Signal') ylabel('ECG') subplot(2,1,2) plot(t,x,'b');hold on; plot(S_t,S_amp,'+r'), grid on; hold on; plot(R_t,R_amp,'+k'); hold on;
Early Myocardial Infarction Detection 86 plot (Q_t, Q_amp, '+g'); hold on; plot (T_peak_t,T_peak, '+y');hold on; plot (TP_t,TP_amp, '+m');hold on; plot (J_t,J_amp,'+c');hold on; plot (K_t,K_amp,'*r');hold on; plot (P_t,P_amp,'*m'); title('Biorthogonal Wavelet Transformed ECG Signal with Q-Peaks (green), R-Peaks (black),S-Peaks (red)') ylabel('ECG+S+R+Q+P+J') hold off; fprintf(1,'\nK> Analysis Complete \n'); %*************************************************************** % End of Code %***************************************************************
Early Myocardial Infarction Detection 87 Appendix B he Graphical user inerface for the software code developed is as follows: function varargout = GUISTchange(varargin) gui_Singleton = 1; gui_State = struct('gui_Name', 'gui_Singleton',
mfilename, ... gui_Singleton, ...
'gui_OpeningFcn', @GUISTchange_OpeningFcn, ... 'gui_OutputFcn',
@GUISTchange_OutputFcn, ...
'gui_LayoutFcn',
[] , ...
'gui_Callback',
[]);
if nargin && ischar(varargin{1}) gui_State.gui_Callback = str2func(varargin{1}); end if nargout [varargout{1:nargout}] = gui_mainfcn(gui_State, varargin{:}); else gui_mainfcn(gui_State, varargin{:});
Early Myocardial Infarction Detection 88 end %*************************************************************** % Get Data & User Inputs %*************************************************************** function edit1_Callback(hObject, eventdata, handles) Fname = get(hObject,'string'); handles.Filename = Fname; guidata(hObject, handles); function edit2_Callback(hObject, eventdata, handles) handles.Filename = ''; handles.Result = ''; handles.output = hObject; guidata(hObject, handles); function varargout = GUISTchange_OutputFcn(hObject, eventdata, handles) varargout{1} = handles.output; Fileloc = 'C:\MATLAB7\Work\'; Headerfile = strcat(handles.Filename,'.hea'); In TXT Format
% Header
Early Myocardial Infarction Detection 89 load (handles.Filename);
% .mat file for Data
%*************************************************************** % Load Header Data %*************************************************************** signalh = fullfile(Fileloc, Headerfile); fid1 = fopen(signalh,'r'); z = fgetl(fid1); A = sscanf(z, '%*s %d %d',[1,2]); nosig = A(1);
% Number Of Signals
sfreq = A(2); clear A; z = fgetl(fid1); A = sscanf(z, '%*s %*d %d %d %d %d',[1,4]); gain = A(1); clear A; S = sfreq*60; counter1=0; counter2=0;
% Integers Per mV
Early Myocardial Infarction Detection 90 counter3=0; for n = 0:4 j = S*n+1:1:S*(n+1); D = val(j); dat = length (D); k = 1:1:dat; D = D(k)/gain; %*************************************************************** %Signal filter and Base line wander correction %*************************************************************** D= transpose (D); axes(handles.axes1) plot (D,'g'); title ('\it{Original Signal for ECG}'); ylabel ('Amplitude in Volts'); xlabel ('# of Samples'); %set(handles.axes1,'XMinorTick','on'); set(handles.axes1,'Color','k');
Early Myocardial Infarction Detection 91 drawnow; windowSize = 5; filsig = filter (ones(1,windowSize)/windowSize,1,D); y = medfilt1(filsig,200); % 1st median filter s1 = y; clear y; y = medfilt1(s1,600); % 2nd median filter D = filsig - y; axes(handles.axes2) plot (D,'g'); title ('\it{Filtered Baseline Wander Corrected Signal for ECG}'); ylabel ('Amplitude in Volts'); xlabel ('# of Samples'); set(handles.axes2,'Color','k'); drawnow; clear s1; clear y;
Early Myocardial Infarction Detection 92 pack; %*************************************************************** % Manipulate Data So We Only Look At What The User Wants %*************************************************************** D = transpose (D); D = cwt (D, 1:4, 'bior2.4'); %Performing Continuous Wavelet Transform using Biorthogonal Wavelet to ECG_1 D = transpose (D); D_1 = D (:,1); D_2 = D (:,2); D_3 = D (:,3); x = D (:,4); clear D; %*************************************************************** % R-Peak Detection %*************************************************************** thresh = 0.6; % create time axis len = length (x);
Early Myocardial Infarction Detection 93 tt = 1/sfreq:1/sfreq:ceil(len/sfreq); t = tt(1:len); max_h = max (x(round(len/4):round(3*len/4)));%segment search area first find the highest bumps in the ECG_1 poss_reg = x>(thresh*max_h); %then build an array of segments to look in left
= find(diff([0 poss_reg'])==1); % remember to zero pad at
start right = find(diff([poss_reg' 0])==-1); % remember to zero pad at end for(i=1:length(left)) [maxval(i) maxloc(i)] = max( x(left(i):right(i)) ); [minval(i) minloc(i)] = min( x(left(i):right(i)) ); maxloc(i) = maxloc(i)-1+left(i); % add offset of present location minloc(i) = minloc(i)-1+left(i); % add offset of present location end R_index = maxloc;
Early Myocardial Infarction Detection 94 R_t
= t(maxloc);
R_amp = maxval; %*************************************************************** % Heart Rate Calculation %*************************************************************** for j = 2:length (R_t) HR(j)= R_t(j)-R_t(j-1); end H_R = 60/(mean (HR)); %*************************************************************** % S-Point Detection %*************************************************************** R_len= length (R_index); for j = 1:R_len IR1 = R_index(j); for i = IR1:IR1+ (round(sfreq*0.03*(H_R/72))) if i == length (x)|i==0 S_index(j)= 1; S_amp(j) = x(1,1);
Early Myocardial Infarction Detection 95 S_t(j) = t(1,1); break end if x(i,1)< x(i+1,1) && x(i,1)< x(i-1,1) S_index(j)= i; S_amp(j) = x(i,1); S_t(j) = t(1,i); break end end end %*************************************************************** % Q-Point Detection %*************************************************************** for j = 1:R_len IR1 = R_index(j); for i = IR1:-1:IR1- (round(sfreq*0.03*(H_R/72))) if i == 0|i==length (x)|i-1==0
Early Myocardial Infarction Detection 96 Q_index(j)= 1; Q_amp(j) = x(1,1); Q_t(j) = t(1,1); break end if x(i,1)< x(i+1,1) && x(i,1)< x(i-1,1) Q_index(j)= i; Q_amp(j) = x(i,1); Q_t(j) = t(1,i); break end end end %*************************************************************** % J-Point Detection %*************************************************************** S_len = length (S_index); for j = 1:S_len
Early Myocardial Infarction Detection 97 IS1= S_index(j); foundj = 0; for i=IS1:IS1+ (round(sfreq*0.03*(H_R/72))) if i==0 J_index(j)=1; J_amp(j)= x(1,1); J_t(j)= t(1,1); foundj = 1; break end if i > length (x) break end if x(i,1)>=0 J_index(j)=i; J_amp(j)= x(i,1); J_t(j)= t(1,i); foundj = 1;
Early Myocardial Infarction Detection 98 break end end if foundj == 0 J_index(j)=1; J_amp(j)= x(1,1); J_t(j)= t(1,1); end end %*************************************************************** % T-Peak Detection %*************************************************************** J_len = length (J_index); for j = 1: J_len P1 = R_index(j)+ (round(sfreq*0.4*(H_R/72))); P2 = J_index (j)+ (round(sfreq*0.08*(H_R/72))); if P1> length (x)|P2> length (x) break
Early Myocardial Infarction Detection 99 end if P1 > P2 [T_peak(j),T_peak_index(j)] = max(x(P2:P1)); T_peak_index(j) = T_peak_index(j)+ P2; else [T_peak(j),T_peak_index(j)] = max(x(P1:P2)); T_peak_index(j) = T_peak_index(j)+ P1; end T_peak_t (j)= t(T_peak_index (j)); end
%*************************************************************** % T-Point (T onset) Detection via T-peak %*************************************************************** T_len = length (T_peak_index); for j = 1:T_len IT1 = T_peak_index(j); for i = IT1:-1:IT1-(round(sfreq*0.035*(H_R/72)))
Early Myocardial Infarction Detection 100 TP_index(j)=i; TP_amp(j)=x(i,1); TP_t(j)=t(1,i); end end %*************************************************************** % K-Point %*************************************************************** Q_len = length (Q_index); for j = 1:Q_len IQ1 = Q_index(j); foundk = 0; for i=IQ1:-1:IQ1- (round(sfreq*0.03*(H_R/72))) if i == 0 K_index(j) = 1; K_amp(j)= x(1,1); K_t(j)=t(1,1); foundk = 1;
Early Myocardial Infarction Detection 101 break end if x(i,1)>=0 K_index(j) = i; K_amp(j)= x(i,1); K_t(j)=t(1,i); foundk = 1; break end end if foundk == 0 K_index(j)=1; K_amp(j)= x(i,1); K_t(j)=t(1,i); end end
Early Myocardial Infarction Detection 102 %*************************************************************** % P-Point Detection via K+80ms %*************************************************************** K_len = length (K_index); for j = 1:K_len IK1 = K_index(j); for i=IK1:-1:IK1- (round(sfreq*0.08*(H_R/72))) if i ==0 P_index(j) = 1; P_amp(j)= x(1,1); P_t(j)=t(1,1); break end P_index(j) = i; P_amp(j)= x(i,1); P_t(j)=t(1,i); end end
Early Myocardial Infarction Detection 103 %*************************************************************** % Calculation of Isoelectric Line %*************************************************************** j = 1:1:K_len; ISO(j) = mean(x(P_index(j):K_index(j))); %*************************************************************** % Calculation of ST-segment %*************************************************************** a = length (J_index); b = length (TP_index); if a==b; j = 1:1:J_len; ST(j) = mean(x(J_index(j):TP_index(j))); end if a>b j = 1:1:b; ST(j) = mean(x(J_index(j):TP_index(j))); end if a= ST(j)+0.0001 && ISO(j)>=ST(j)0.0001)|ISO(j)==ST(j) %fprintf(1,'\nK>No MI\n'); counter2=counter2+1; else counter3=counter3+1; end
Early Myocardial Infarction Detection 105 end end if a=ST(j)+0.0001 && ISO(j)>=ST(j)0.0001)|ISO(j)==ST(j) counter2=counter2+1; else counter3=counter3+1; end end end
if a>b for j=1:b counter1=counter1+1;
Early Myocardial Infarction Detection 106 if (ISO(j)>=ST(j)+0.0001 && ISO(j)>=ST(j)0.0001)|ISO(j)==ST(j) counter2=counter2+1; else counter3=counter3+1; end end end clear ISO; clear ST; if counter3/counter1>=0.90 fprintf(1,'\nK>WARNING: Possible MI\n'); else fprintf(1,'\nK>No MI\n'); end axes(handles.axes3) plot(t,x,'g');hold on; plot(S_t,S_amp,'+r'), hold on;
Early Myocardial Infarction Detection 107 plot(R_t,R_amp,'+w'); hold on; plot (Q_t, Q_amp, '+b'); hold on; plot (T_peak_t,T_peak, '+y');hold on; plot (TP_t,TP_amp, '+m');hold on; plot (J_t,J_amp,'+c');hold on; plot (K_t,K_amp,'Marker','+','color',[1,0.4,0.6]);hold on; plot (P_t,P_amp,'Marker','+','color',[0.4,0,0.6]); title('Biorthogonal Wavelet Transformed ECG Signal with Q-Peaks (green), R-Peaks (black),S-Peaks (red)') ylabel('ECG+S') hold off; ylabel ('Amplitude in Volts'); xlabel ('Time in seconds'); set(handles.axes3,'Color','k'); drawnow; end if counter3/counter1>=0.90 fprintf(1,'\nK>WARNING: Call 9-1-1 & Get Help\n');
Early Myocardial Infarction Detection 108 handles.Result = 'WARNING: Call 9-1-1 & Get Help'; else fprintf(1,'\nK>No ST-Changes Detected\n'); handles.Result = 'No ST-Changes Detected'; end guidata(hObject, handles); set(handles.edit2,'String',handles.Result); function edit1_CreateFcn(hObject, eventdata, handles) if ispc set(hObject,'BackgroundColor','white'); else set(hObject,'BackgroundColor',get(0,'defaultUicontrolBackgroundC olor')); end function edit2_CreateFcn(hObject, eventdata, handles) if ispc set(hObject,'BackgroundColor','White');
Early Myocardial Infarction Detection 109 else set(hObject,'BackgroundColor',get(0,'defaultUicontrolBackgroundC olor')); end function figure1_ResizeFcn(hObject, eventdata, handles) %*************************************************************** % End Of Code %***************************************************************
Early Myocardial Infarction Detection 110 Appendix C Testing Results Dataset Name
Sample #
MIT-BIH ST Change Database 5 minutes of data tested
327m 325m 324m 323m 322m 321m 320m 319m 318m 317m 316m 315m 314m 313m 312m 311m 310m e0103m e0105m e0107m e0113m e0111m e0119m e0121m e0123m e0125m e0127m e0133m e0139m e0147m e0155m 118e00m 118e06m 118e12m 118e18m 118e24m s20021m s20111m
European ST Database 5 minutes of data tested
MIT-BIH Noise Stress Database 5 minutes of data tested
Longterm ST Database 5 minutes of data tested
MI
No MI 268 375 248 375 254 378 0 133 565 144 115 379 181 105 375 83 0 60 0 0 305 310 240 302 299 349 293 265 156 265 203 0 0 0 0 0 431 456
0 0 62 85 241 0 411 407 169 213 425 0 252 345 0 308 514 245 273 330 0 0 60 77 71 0 74 0 234 0 148 361 361 361 361 361 0 0
Total # of Signals 268 375 319 460 495 378 411 540 734 357 540 379 433 450 375 391 514 305 273 330 305 310 300 379 370 349 367 265 390 265 351 361 361 361 361 361 431 456
Early Myocardial Infarction Detection 111 s20141m s20651 s20231m s20081m MIT-BIH Normal Sinus Rhythm Database 5 minutes of data tested
590 530 511 429
0 0 0 89
590 530 511 518
16265m
0
298
298
16272m 18184m 19093m 19140m 19830m 16273m 16420m 16483m 16773m 16786m 17453m 18177m
0 292 234 196 0 0 273 290 88 0 171 0
328 179 68 248 156 279 180 57 324 318 248 175
328 471 302 444 156 279 453 347 412 318 419 175
Results with 100% accuracy for MI Results with 100% accuracy for Normal ECG
Early Myocardial Infarction Detection 112 Appendix D Receiver Operating Characteristics Curve Datasheet Sample # 327m 325m 324m 323m 322m 321m 320m 319m 318m 317m 316m 315m 314m 313m 312m 311m 310m e0103m e0105m e0107m e0113m e0111m e0119m e0121m e0123m e0125m e0127m e0133m e0139m e0147m e0155m 118e00m 118e06m 118e12m 118e18m 118e24m
MI
NMI 268 375 248 375 254 378 0 133 565 144 115 379 181 105 375 83 0 60 0 0 305 310 240 302 299 349 293 265 156 265 203
0 0 62 85 241 0 411 407 169 213 425 0 252 345 0 308 514 245 273 330 0 0 60 77 71 0 74 0 234 0 148
0 0 0 0 0
361 361 361 361 361
Percentage Percentage Total Correct Wrong Databse Name 268 100 0 MIT-BIH ST Change 375 100 0 319 77.74294671 19.43573668 460 81.52173913 18.47826087 495 51.31313131 48.68686869 378 100 0 411 0 100 540 24.62962963 75.37037037 734 76.97547684 23.02452316 357 40.33613445 59.66386555 540 21.2962963 78.7037037 379 100 0 433 41.80138568 58.19861432 450 23.33333333 76.66666667 375 100 0 391 21.22762148 78.77237852 514 0 100 305 19.67213115 80.32786885 European ST 273 0 100 330 0 100 305 100 0 310 100 0 300 80 20 379 79.68337731 20.31662269 370 80.81081081 19.18918919 349 100 0 367 79.83651226 20.16348774 265 100 0 390 40 60 265 100 0 351 57.83475783 42.16524217 MIT-BIH Noise Stress 361 0 100 Test 361 0 100 361 0 100 361 0 100 361 0 100
Early Myocardial Infarction Detection 113 s20021m s20111m s20141m s20651 s20231m s20081m 16265m 16272m 18184m 19093m 19140m 19830m 16273m 16420m 16483m 16773m 16786m 17453m 18177m
431 456 590 530 511 429 0 0 292 234 196 0 0 273 290 88 0 171 0
Diagnostic Level 90% 80% 70% 50% 40%
0 0 0 0 0 89 298 328 179 68 248 156 279 180 57 324 318 248 175
431 456 590 530 511 518 298 328 471 302 444 156 279 453 347 412 318 419 175
100 100 100 100 100 82.81853282 0 0 61.99575372 77.48344371 44.14414414 0 0 60.26490066 83.57348703 21.3592233 0 40.81145585 0
0 Longterm ST change 0 0 0 0 17.18146718 100 Normal Sinus Rhythm 100 38.00424628 22.51655629 55.85585586 100 100 39.73509934 16.42651297 78.6407767 100 59.18854415 100
