PCI procedure
MELKER HAGELBERG
Master’s Thesis in Engineering Physics Examiner: Mats Danielsson
TRITA-FYS 2016:47 ISSN 0280-316X ISRN KTH/FYS/–16:47—SE
Introduction: One of the most common causes of death in Sweden is cardiovascular disease. The disease is mainly caused by reduction in size of the arterial vessels in the heart which disturb the blood flow supplying oxygen to the heart. Narrowed blood vessels are often treated with a method called Percutaneous Coronary Intervention (PCI). When performing a PCI procedure, a plastic tube called a catheter is used.
This catheter is often equipped with a balloon in the end, that can be inflated by the doctor at the desired location in the vessel. The catheter is guided through the arteries up to the narrowed vessel and the narrowing is opened by inflating the balloon. The guidance of the catheter is performed using X-rays. The X-rays used in the examination may lead to devastating injuries for both the patient and the staff who work in the examination room if they are not accurately protected.
Aim: The aim of this thesis is to find ways to reduce radiation dose during an interventional procedure. The improvements were found by investigating the workflow and the technical factors that the clinic was working with when this thesis was initiated.
Method: The thesis is divided into three different parts; a littera- ture study, a pre-study and a pilot study. The literature study had the aim to find out which factors affect the dose during a PCI procedure. In the pre-study a number of parameters were tested on the basis of both image quality and dose. The goal of the pre-study was to investigate suitable parameters that could be changed and used in the pilot study.
In the pilot study a new dose protocol was implemented in the ma- chine and used when examining the patients. A total of 70 patients were evaluated based on procedure type, physiological parameters of the pa- tient and the dose delivered during the procedure. These parameters were then compared with retrospective data. Both the pre-study and the pilot study were performed at Vrinnevi Hospital in collaboration with GE Healthcare.
Conclusion: The results of the pilot study showed no significant differences between retrospective data and the data received after im- plementation of the new protocol. However significant results were not expected due to the small sample size. This work can be used as a proposal for a larger pivotal study and demonstrates the importance of constantly working to minimize the dose delivered during a PCI exam- ination.
Bakgrund: En av de vanligaste dödsorsakerna i Sverige är kar- diovaskulära sjukdomar. Dessa uppkommer ofta till grund av en re- duktion av framkomligheten för blod i artärerna som syresätter hjärtat.
Idag använder man till stor del en metod kallad PCI för behandling av förträngda kärl. När man utför en PCI undersökning utnyttjar man ett plaströr som kallas kateter. Denna kateter har i ena änden en liten bal- long som kan blåsas upp av doktorn vid önskat ställe. Katetern guidas genom artärerna upp till det förträngda kärlet som man sedan utvidgar genom att blåsa upp ballongen. Guidningen av kateter utförs med hjälp av röntgenstrålar. Röntgenstrålningen som används under undersöknin- gen kan leda till skador för både patienten och personalen som arbetar i undersökningsrummet om de ej är medvetna och skyddar sig ordentligt mot strålningen. [1]
Mål: Målet med detta arbete är att hitta sätt att minimera den stråldos som levereras från röntgenutrustningen under en interventionell undersökning. Förbättringarna hittades genom att undersöka arbetsflö- den och de tekniska inställningar hos maskinen.
Metod: Detta arbete är indelat i flera delar. En literaturstudie med målet att ta reda på vilka faktorer som påverkar stråldosen vid en undersökning, en förstudie där ett antal parametrar testades utifrån både bildkvalitét och dos samt en pilotstudie. Målet med förstudien var att undersöka lämpliga parametrar som kan implemeteras i pilotstudien.
I pilotstudien implementerades ett nytt dosprotokoll i röntgenmaskinen och detta protokoll användes sedan vid undersökningar av 70 patienter.
Patienterna utväderades utifrån procedur typ, fysologiska parametrar och den dos som levererades under undersökningen. Dessa parame- trar jämnfördes sedan med retrospektiv data. Både förstudien och den kliniska studien utfördes på Vrinnevi Sjukhuset i sammarbete med GE Hhealthcare.
Slutsats: Resultatet av pilotstudien visar inga signifikanta skill- nader mot den insamlade retrospektiva datan. En signifikant skillnad var ej förväntad då mängden datapunkter var liten. Detta arbetet kan användas som ett förslag till en större mer omfattande studie och påvisar vikten av att ständigt arbeta för att minimera den dos som levereras vid en PCI undersökning.
This Thesis was carried out and performed in collaboration with GE Healthcare Sweden. It has been a pleasure working with people and technology that are the best there is in this field at this time.
First of all, I want to thank my supervisor Magnus Skytt for the support through out this project.
Secondly, I want to thank all the people at Vrinnevi Hospital for their time, patience and help.
A special thanks to Andreas Forsberg (Sunderby Hospital) for help- ing me with the set-up of my pre-study and Markus Kaila (GE health- care) for helping me with the scientific review of this thesis.
Finally, I want to thank to my family for all the help and support through these years.
CABG - Coronary Artery Bypass Graft Cine - Cineangiography
CTO - Chronic Total Occlusion DQE - Detector Quantum Efficiency DAP - Dose Area Product
ECG - Electrocardiography FID - Focus to Image Distances FOV - Field Of View
fps - frames per second
GFR - Glomerular Filtration Rate IRP - International Reference Point IQ - Image Quality
LAD - Left Anterior Descending LCA - Left Coronary Artery LCX - Left Circumflex Artery LDL - Low Density Lipoprotein MI - Myocardial Infarct
OID - Object to Image Distance
PCI - Percutaneous Coronary Intervention PMMA - PolyMethyl MethacrylAte RCA - Right Coronary Artery RDL - Receptor Dose Limited ROI - Region Of Interest SID - Source to Image Distance SNR - Signal Noise Ratio
1 Introduction 1
1.1 Thesis objective and definition . . . 2
1.2 Limitations . . . 3
2 Theoretical background 5 2.1 Anatomy of the heart and heart diseases . . . 5
2.1.1 Atherosclerosis . . . 6
2.1.2 Symptoms . . . 7
2.2 Catheterization . . . 7
2.2.1 Angiography . . . 7
2.2.2 Percutaneous Coronary Intervention (PCI) . . . 10
2.2.3 Contrast solution . . . 12
2.3 X-ray production and its interaction with matter . . . 12
2.3.1 Interaction of x-rays with matter . . . 14
2.4 Equipment . . . 14
2.4.1 Dose . . . 16
2.4.2 Dose settings . . . 16
2.5 Diagnostic value . . . 17
3 Method 19 3.1 Workflow - Vrinnevi Hospital . . . 19
3.1.1 Pre-operation . . . 19
3.1.2 Operation . . . 19
3.1.3 Post-operation . . . 21
3.2 Literature study . . . 22
3.2.1 Operator dependent . . . 22
3.2.2 Patient dependent . . . 23
3.2.3 Technically dependent . . . 24
3.3 Pre-study . . . 24
3.3.1 Basic elements of the pre-study . . . 24
3.3.2 Patient thickness . . . 26
3.3.3 Field Of View (FOV) . . . 27
3.3.4 Object to Image Distance (OID) . . . 28
3.3.7 Dose protocol . . . 30
3.4 Pilot Study . . . 31
3.4.1 Statistics . . . 31
3.4.2 Radiation dose measurements . . . 32
3.4.3 Baseline comparison . . . 32
4 Results 35 4.1 Pre-study . . . 35
4.1.1 Patient thickness . . . 35
4.1.2 Field Of View (FOV) . . . 36
4.1.3 Object to Image Distance (OID) . . . 37
4.1.4 Frame rate . . . 38
4.1.5 Collimation . . . 39
4.1.6 Dose protocol . . . 40
4.2 Pilot Study . . . 42
4.2.1 Angiography . . . 42
4.2.2 PCI . . . 42
5 Discussion 43 5.1 Pre-study . . . 43
5.1.1 Patient thickness . . . 43
5.1.2 Field Of View (FOV) . . . 43
5.1.3 Object to the Image Distance (OID) . . . 43
5.1.4 Frame rate . . . 44
5.1.5 Collimation . . . 44
5.1.6 Dose protocol . . . 44
5.2 Pilot Study . . . 45
5.2.1 Angiography . . . 45
5.2.2 PCI . . . 45
5.2.3 Comparison with a similar study . . . 46
5.2.4 Recommendation for pivotal trials . . . 47
6 Conclusions 49
Bibliography 51
Appendices 53
A 55
Introduction
In year 2013, 37% of the people that died in Sweden, died because of cardiovascular diseases, making it the most common cause of death [2]. Cardiac catheterization is used in medicine to diagnose and treat heart conditions caused by stenosis. Stenosis is a condition where the blood vessel is narrowed, reducing the bloodflow within the vessel. Cardiac catheterization is performed using a catheter, which is a thin flexible tube that can be fitted into the blood vessel. The entrance point is usually the groin or the arm. The catheter can be used to find lesions in the heart and in its surrounding vessels (coronary arteries). The catheter can be equipped with a balloon in the end of it tip, which is used to expand the size of the vessel from the inside. By expanding the balloon the stenosis is pressed into the vessel wall and the blood flow is re-established. This kind of procedure is called angioplasty and is one of many techniques that is included in the term Percutaneous Coronary Intervention (PCI).
Since the first catheterization in year 1711, the technique has developed progres- sively and most of the physiological information of the cardiovascular system have been investigated using catheters [3]. One of the big additions to the technique is the introduction of stent covered balloons. A stent is a metal mesh that is used to re- tain the normal vessel size after expanding it with a balloon catheter [3]. Currently, catheterization allows accurate diagnosis of almost all major cardiac diseases and the safe treatment of multiple cardiac diseases, including coronary artery disease.
The continuously increasing amount of people suffering from coronary diseases all over the world make PCI and the surgical treatment of coronary disease one of the most important developments in medicine [3].
An important factor to reduce mortality is to diagnose and treat the stenosis in an early stage. Diagnosis of stenosis in the coronary artery is performed using angiography and the treatment is often performed using PCI [4]. In year 2014, 40 000 angiographies and 25 000 PCIs were performed in Sweden [5]. During an interventional procedure both cinegraphy and fluoroscopy, which are two types of x-ray delivery techniques, are used to guide the catheters in the blood vessel to the place of the stenosis. The usage of x-rays during the procedures result in
high radiation exposure to the patient and the clinicians which increase the risk of sustainable damage [6].
When performing angiography and PCI the radiation dose of the patient and the clinical staff are influenced by multiple parameters including fluoroscopic time, frame rate, number of images and the usage of radiation protection. The reduction of radiation delivered by the system will minimize the total radiation dose but also the secondary radiation that is scattered when hitting the initial target. The secoundary radiation increases the image noise and reduces the image quality [4].
When working in the field of cardiovascular interventions reducing the dose and keeping the diagnostic value in the images is the main focus. The trend shows that the numbers of PCIs increase every year [5] which confirm the importance of minimizing the dose delivered from the system to the patient and care giving staff.
1.1 Thesis objective and definition
The aim of this thesis is to investigate the parameters that affect radiation dose and find improvements within the workflow, settings or set-up to reduce the dose delivered by the system during a procedure, without loosing the diagnostic value of the images. The thesis consists of the following parts:
• A literature study combined with clinical visits to see how the clinic is working and learn how a procedure is performed.
• A pre-study, where a couple of factors from the literature study further in- vestigated, based on radiation dose and image quality. This step is done to make sure that the change performed is safe and that the dose levels are lower than the original settings. The dose and the image quality are plotted and evaluated before starting the pilot study in the real clinical environment.
• A pilot study is performed to investigate how the changes fits into the real clinical environment. A statistical test is executed to compare retrospective data with the data collected during the time for the pilot study. The categories analysed and compared in the pilot study are:
– Age – Gender
– Body Mass Index (BMI)
– Prior Coronary Artery Bypass Surgery (CABG) – Prior Percutaneous Coronary Intervention (PCI) – Access point
– Number of stenosis – Number of stents
– Usage of addition techniques
– Chronic Total Occlusion (CTO) – Contrast
– Fluoroscopic time
– Dose Area Product (DAP)
1.2 Limitations
This study is part of a larger effort to test the feasibility of implementing different dose reductive factors and see their effect when performing a PCI. The sample size of the pilot trial was small and no significant results were expected to be found.
However, rather then looking for significant results, the aim was to implement im- provements and to recommend a set-up for a pivotal study.
The data collection is only gathered from Seldinger lab 7 at Vrinnevi Hospital and only analysed from one doctor which limit the number of patients that can be evaluated. Collection of data was performed between 2015-11-02 and 2015-12-02.
Theoretical background
2.1 Anatomy of the heart and heart diseases
The heart is supplied with blood from the right and left coronary arteries, both originating from the aortic root. The Right Coronary Artery (RCA) supplies the right ventricle and atrium with blood. The Left Coronary Artery (LCA) is divided near its entrance point into two big branches: the Left Circumflex Artery (LCX) and the Left Anterior Descending (LAD). LCX goes around the heart to the left and supplies the left side of the wall in the left ventricle. LAD supplies interventricular septum (wall between the left and the right ventricle) and left ventricle [7].
Figure 2.1: The anatomy of the heart [8].
Chest pain is one of the main signs of heart disease. The pain typically radiates or is isolated to the neck, jaw, shoulder or the left arm area and often increases with the severity of the disease. Other symptoms that may appear are shortness of breath, sweating, nausea and vomiting [7]. The cause of heart disease depends
on multiple factors, but the most usual cause is the production of atherosclerosis.
Atherosclerosis refers to the production of plaque in the wall of the arteries. The growth of the plaque leads to reduction of the lumen (the area inside the vessel) size, which decreases the blood flow in the coronary arteries. The phenomenon when the lumen size is reduced by plaque is called stenosis. Stenosis is easiest to diagnose using angiography and, if needed, treated using PCI [7].
2.1.1 Atherosclerosis
Atherosclerosis is developed as an effect of an inflammatory response that later will produce plaque within the arteries. The inflammation is initiated when Low Density Lipoprotein (LDL), a type of cholesterol, advances over the vessel wall from the bloodstream into the intima, which is the area on the other side of the endothelium (see Figure 2.2). Endothelium are the cells producing the inner wall of the vessel, separating the vessel lumen and the intima. Within the intima the cholesterol is oxidised [9].
The development and growth of atherosclerosis caused by an injury or dysfunc- tion leads to an over expression of binding cites for lymphocytes (white blood cells) in the endothelial wall, making it possible for lymphocytes to enter the intima [10].
Figure 2.2: The structure of the coronary vessels and a vessel narrowed due to atherosclerosis [11].
The presence of lymphocytes in the intima will release cytokine. Cytokine is a protein that transforms the lymphocytes to a type of white blood cells that ingest the oxidized lipid particles with the goal to eliminate the LDLs presence in the intima. Because of the high concentration of oxidized cholesterol, the lymphocytes are transformed into dead lipid-filled foam cells that are left behind. The cytokine will also make the smooth-muscle cells grow and produce a protecting cap for the the foam cells [10].
The change in the endothelial permeability in combination with the smooth- muscle cells in the intima will increase the inflow of cholesterol into the arterial wall
and produce even more foam cells. The foam cells will eventually die, leaving the lipid rich necrotic core of the plaque behind. The fibrous cap covering the necrotic core is very vulnerable and could be disrupted leading to a total block further down in the blood stream. This is called a Chronic Total Occlusion(CTO) [10] [12].
2.1.2 Symptoms
Chest pain with different severity is the usual effect of a narrowing of one or more of the coronary arteries. The extent of the damage varies depending on the number of vessels affected and the size of blockage.
Heart diseases are separated into two different categories, Myocardial ischemia and Myocardial infarct (heart attack). The separation depends on the severity of the blockage, leading to reduced oxidization of the heart muscle. Myocardial ischemia is a less severe version of heart disease with the possibility of total recovery of the heart muscle after treatment. Myocardial infarct is defined as death of the heart muscle and is the worst type of heart disease and may lead to sustainable damages of the heart muscle if not treated quickly. The worst type of myocardial infarction is caused by a total occlusion in one of the main coronary arteries. A total occlusion needs to be treated as quickly as possible to reduce the risk of permanent damage of the heart [12] [13].
2.2 Catheterization
Catheterization is a combination of a hemodynamic and angiographic procedure with diagnostic and therapeutic purpose. The aim of cardiac catheterization in a patient with coronary disease is to identify the suspected abnormalities in the vessel using angiography and restore the arterial blood flow through PCI [1].
2.2.1 Angiography
Coronary angiography is one of the best ways to investigate important characteris- tics of the coronary vessels such as anatomy, formation and pathology. The main advantage of using angiography as a diagnostic procedure when investigating the coronary vessel is the possibility to shift between diagnostic and therapy, using PCI as a therapeutic method through the same access site [1].
An angiography is usually performed using the percutaneous approach (from the femoral or radial artery, see Figure 2.3). 80% of the angiographies performed in Sweden use the percutaneous approach through the radial artery due to the re- duced risk of bleeding compared to examination done through the femoral artery [5].
Figure 2.3: The left arrow show the access site through the radial artery. The right arrow show the access site to the femoral artery [14].
Percutaneous approach through the radial artery
The aorta gives rise to the subclavian artery that in the end merges into the radial artery. After finding the puncture site the procedure is initiated by injecting local anaesthesia into the skin of the patient’s wrist. The modified Seldinger technique is used for the introduction of guidewires and catheters. The procedure is performed using the following steps:
1. The modified Seldinger technique is initiated by cutting a small hole with a scalpel and then inserting a Seldinger needle into the radial artery (see Figure 2.4) [1].
Figure 2.4: Three Seldinger needles surrounded by a guidewire [1].
2. A guidewire is then introduced through the needle (see Figure 2.4). The modern guidewires have a soft tip, the ability to control the torque of the tip and are visible on the x-ray. The guidewire is constructed with a solid core to give stability and becomes thinner in its distal part. The core is then covered with a spring coil that bend the front tip of the guidewire when pulling the spring in the end part of the wire [1].
3. The needle is then removed and a hydrophilic introducer (see Figure 2.5) is inserted by sliding it over the guidewire. The introducer has a valve that stops blood from leaking up through its pipe. The introducer also has a side-arm connector that controls the bleeding around the catheter shaft and used as a port for injecting extra intravenous fluids or drugs (contrast solution) during the procedure. The introducer also works as a channel into the artery for different kinds of catheters and guidewires [1].
Figure 2.5: Three different types of introducers [1].
4. The guidewire is inserted through the introducer and guided up through the aortic arch into the beginning of the coronary arteries [1].
5. The catheter (see Figure 2.6) is then advanced into the aortic root over a guidewire and guided using fluoroscopy. When the catheter have reached the beginning of the coronary arteries the guidewire is removed and the catheter is manoeuvred into the coronary arteries. When the catheter is in the desired place in the beginning of the coronary arteries, a continuous stream of con- trast is injected through the side arm of the introducer while the patient is irradiated and a coronary angiography is performed [1] [7] [15].
Figure 2.6: Five different types of catheters. The two first from the left are diag- nostics catheters used for angiography. The three from the right have a balloon attached in the end of the catheter that is used when performing a PCI [1].
2.2.2 Percutaneous Coronary Intervention (PCI)
PCI is a broad designation including devices and techniques used to solve prob- lems relating to elastic recoil, dissection and restenosis of a vessel. PCI includes different forms of catheter based techniques where the main procedure of coronary intervention is PCI and angioplasty [7].
Before starting the procedure an angiography is produced to evaluate the present condition of the vessel, find the lesions and choose the projection that gives the best image of the stenosis. A guidewire is then introduced in the catheter and used during the angiography. The guidewire is advanced across the lesion and with a small injection of contrast material and fluoroscopy is the vessel imaged to confirm the wire position. When the guidewire is in the right place, a balloon catheter is slided over it to the position across the stenosis (see Figure 2.7) and the balloon is inflated [7].
Figure 2.7: A) The dilation balloon positioned in the right place. B)A dilated balloon. C) The result after balloon PCI [7].
The important features of the dilatation balloon catheter are how easy it passes curved segments and the ability to be inflated to an exact defined size at a certain pressure [7]. An overextended balloon can lead to an overstretched vessel or that the balloon bursts inside of the vessel. The balloon is filled with a diluted contrast solution to reduce the complications with air in the blood stream if the balloon bursts [7].
Once the dilatation catheter has been positioned within the target stenosis, the balloon is inflated using a screw-powered hand held inflation device [7]. When a stent should be placed, the vessel often is pre-dilated with a balloon that is slightly smaller than the reference vessel and approximately the same length as the lesion.
A stent is a metallic mesh that is deployed in a diseased vessel to obtain or maintain the original luminal size and reduce restenosis (see Figure 2.8). After the stent is successfully placed and checked using angiography, the balloon and catheters are removed [7].
Figure 2.8: Stent used for coronary stenting [7].
The angioplasty will redistribute the plaque within the vessel. The main impact will be over stretching the vessel, making the lumen and the outer diameter of the vessel bigger. Theoretically, a total recovery of the vessel lumen size is possible after a balloon dilation but in reality, due to elastic recoil, the recovery will only be 70%.
Stenting on the other hand eliminates the elastic recoil and the dissection (plaque fracture) which minimize the residual stenosis. The stents used are often covered in drugs (drug eluting stents). The drugs reduce the inflammation that can appear after stent placement [7].
2.2.3 Contrast solution
An angiography and PCI are performed by injecting a contrast solution directly into the coronary arteries and while radiating the patient using x-rays [1].
In angiography, a positive agent is used that makes the images darker by ab- sorbing x-rays, which makes it easier to see any abnormalities of the vessel. The positive contrast agent is often produced from organic iodine and used in many x-ray applications. The contrast solution needs to be injected with a sufficient rate and volume to obtain an effective replacement of the blood in the vessels. The operator will flush the vessel with enough fluid to get a continuous flow back into the aortic root [7] [12].
The patient can sometimes react negatively to the contrast agent. This reac- tion could come from an allergic reaction or be related to the chemical property of the agent. The most usual side effects are headache, nausea, vomiting and muscu- loskeletal pain. More severe side effects like renal dysfunction and renal disease can also appear [7] [12].
2.3 X-ray production and its interaction with matter
X-rays are produced when electrons with high energy are stopped during the impact with a metal target. Most of the kinetic energy of the electrons is transformed into heat (99%), but x-rays are also produced in this process (1%) [16].
The X-ray tube (see Figure 2.9) consists of a negative cathode (filament) and a positive anode (metal target). The filament is heated by passing electrical current through it. When reaching high temperatures it starts to incandescent. In this state it repels free electron from the filament and bombards the targeting anode with a kinetic energy equal to the voltage between the cathode and the anode [16].
Figure 2.9: Structure of an X-ray tube [7].
The production of x-rays is a three-step process (see Figure 2.10) that occurs in the anode. The first step starts when a high energetic free electrons hit the anode and collides with an electron in the K or L-shell in the atom of the metal. The collision produces an electron hole in the K-shell. This hole is filled by an electron
that falls down from the L-shell. During this process a photon (called Kα photon) is released with the energy equal to [16]:
EP hoton= EK−shell− EL−shell(1)
Figure 2.10: Three step process for production of X-ray [17].
X-rays produced in this way consists of a continuous energy spectrum of bremsstrahlung (see Figure 2.11) with specific peaks in the spectrum. The two main peaks are produced from the three step process and are called K-edge and L-edge. Increas- ing the voltage level (kV) of the tube will affect the number of photons produced and its maximum energies. Increasing the current (mA) will increase the amount of bremsstrahlung and characteristic radiation proportionally. X-ray can interact with matter in three different ways [16]:
1. The photon can be absorbed by photoelectric effect 2. The photon can be scattered bu compton scattering 3. The photon can be transmitted through the material
Figure 2.11: The energy spectrum produced by a X-ray tube [17].
The x-ray image is produced by the transmitted photons that are absorbed by the detector. The sum of the absorbed and scattered photons are represented as attenuation by matter in the image. The parameter that quantifies the attenuating
properties of the material is linear attenuation coefficient and it measures the prob- ability that a photon will interact, per unit length on its path, when travelling in a specific material. By dividing the linear coefficient by the density of the material the mass attenuation coefficient µ is obtained. The attenuation is then calculated from the equation [16]:
I = I0e−µd
where I is the detected intensity, I0 is the initial intensity from the tube and d is the distance that the photons are travelling through the material. Calculating the µ makes it possible to identify the imaged tissue and separate different tissue from each other [16].
2.3.1 Interaction of x-rays with matter
Radiation can be produced naturally or by humans. The biggest source of human- made exposure is medical imaging. When x-rays are absorbed or scattered within the human body radiation injuries can appear due to stochastic or deterministic effects. Stochastic effect is the process when a single photon interacts with the DNA of a single cell and produces an unrepairable injury that kills or mutates the cell into a cancer cell. The risk that the stochastic effect appears is increased linearly with dose [7].
Deterministic effects occur when many cells are damaged and visible injuries appear. The injuries appear due to the massive cell death. Skin injuries are the most common deterministic effect when performing interventional procedures, but it is very seldom that the dose level reach the threshold for deterministic effects [7].
2.4 Equipment
The main function of the x-ray cinefluorograpic system is to produce a collimated x-ray beam that is radiated through the patient and to produce images that help the procedure. X-ray cinefluorographic units have the possibility to operate in both fluoroscopy and cineangiography(cine) mode. The big difference between the two modes is the image quality and level of dose delivered to the patient.
Fluoroscopy produces a real-time x-ray image with an image quality optimal for guidance performance. The system used today have a pulsed fluoroscopy, meaning that the x-ray beam is only turned on during a short period of time to decrease the effect of the motion artefacts from the heart, but also to reduce radiation dose [12].
The cineangiography(cine) mode produce images with high image quality for single-frame viewing. Higher doses are delivered to the patient because of demands of reduction of noise and higher image quality. Many cinefluorographic units are calibrated in such a way that the per-frame dose for cine is around 10 times higher than the one for fluoroscopy [12].
The equipment used in this thesis is a GE Innova IGS 520 (GE Healthcare Buc, France). The system gantry (see Figure 2.12) is designed with a floor mounted
L-arm and on the top of the L-arm, the C-arm is attached. At the lower end of the C-arm the X-ray tube is fixed and on the top the flat panel a solid-state digital detector is mounted [18].
The angle of the C-arm in combination with radiation is called projection. The projection is named after orientation of the C-arm. If the patient is laying on the back with the head towards the gantry and the detector is perpendicular over the patient then the projection is called Anterior-Posterior (AP). The projection when the detector is tilted towards the feet is called Caudal, towards the head Cranial, to the patients right Right Anterior Oblique (RAO) and to the patients left Left Anterior Oblique (LAO) [7].
Figure 2.12: The equipment used in this study. The C-arm is to the left with the detector on top and the x-ray tube is on the bottom [19].
The x-ray tube has two filaments that give different focal spot sizes [19]. The smaller is used for fluoroscopy and the larger is used for the higher demanding cine imaging [7]. The tube has a rotating anode with a diameter of 160 mm brazed graphite and an anode with a target angle of 11 degrees [19]. The rotation of the anode is driven by an induction motor and the heat produced at the anode surface is transferred to the housing where it is cooled down by cold water.
The annulus of the target is tilted and the heat is spread out on a larger area of material due to the rotation. The rotation reduces the focal spot and beam angle.
The reduced focal spot and beam angle improves the quality of the x-ray beam [16].
The x-rays leaving the tube will be filtered before interacting with the patient.
Some filters in the system are fixed but additional filtration is added depending on the thickness of the patient [16].
Multi-leaf collimators are also used to reduce unnecessary radiation to the pa- tient and scattered radiation to the clinicians. The radiation goes through the pa- tient and before registration in the detector it passes through an anti-scattered grid.
The anti-scatter grid reduces noise in the image by absorbing scattered radiation produced in the patient just before the photons are detected [16].
The digital detector consists of a cesium iodide (CsI) scintillator that absorbs the the x-ray photons and converts them into light. The light is then converted into electrical charge by amorphous silicon photodiode. The electrical charge of each pixel is turned into digital data that is processed to achieve an image. The detector is 20.5x20.5 cm and consists of 1024x1024 photodiode elements. One photodiode
represents one pixel in the image [18]. In the end of the imaging chain the image is converted into a format called DICOM (Digital Imaging and COmmunication in Medicine) and then stored for later usage in a system called PACS (Picture Archiving Communication System) [12].
2.4.1 Dose
The x-rays that are produced in the tube and interact with the patient will transfer their energy into the patient’s tissue. The radiation going into the patient at a point in space is called Exposure (Air kerma) and has the unit Gray [Gy]. Air kerma is a dosimetric unit that explain how much energy has been released into the medium (dose delivered to air). Exposure by itself does not explain anything about the energy delivered to a patient or the biological effects that can occur due to the absorbed radiation. The highest dose during a procedure is often measured at skin level (skin dose) where the beam enters (entrance dose). At this point the low energetic photons are stopped and high energetic photons are backscattered within the patient. The highest dose delivered to the skin is called peak skin dose [7] [20].
The two most important places to measure dose are the entrance dose and the receptor dose. The entrance dose will give information about the risk of radiation to the patient and the receptor dose will give information about the number of photons that are available for image formation and also determine image noise [7].
The entrance dose is measured in various ways in the fluoroscopic system and is used as an approximation of the actual skin dose. The easiest estimation is the fluoroscopic time but it only considers fluoroscopy. More modern fluoroscopic systems estimate the Dose Area Product (DAP), together with the fluoroscopic- time. DAP is the product between the irradiated field size and the incident radiation (Gycm2) [7]. The DAP and the fluoroscopy time is measured by the system and showed on the monitor [18].
2.4.2 Dose settings
The Innova IGS 520 uses an automatic exposure function called AutoEx. Au- toEx uses the peak kilovoltage (kVp), milliampere (mA), pulse-width, Field Of View (FOV), spectral filtration and average brightness in the image to optimize the settings to obtain the same image quality even if the gantry angles or FOV are changed [18].
In the system it is possible to choose between different protocols that are op- timized for different procedures. The operator can change the protocol together with the frame rate, FOV and the dose detail from the panel on the table. Within the protocols the AutoEx preference settings can be changed between different dose levels and settings that focus on high contrast, lower noise in the image or low radiation dose [18].
Dose Reduction Strategy is a parameter that can be changed to obtain higher dose reduction or focusing on increasing the image quality. The different dose
settings that can be adjusted is presented in the list below [18]:
• Fluoroscopy can be performed at the frame rate: 3.75, 7.5, 15, 30 frames per second and cine can be performed using 15 and 30 frames per second [18].
• The Dose-Details can either be set to low or normal. Changing the dose from normal to low will change the dose rate approximately by 50% [18].
• The operator can change the FOV settings from 20, 17, 15 to 12 cm from the table side [18].
• There are five different AutoEx Preference settings: IQ plus, IQ Standard, Smart IQ, RDL (Receptor Dose Limited) Plus and RDL Standard. IQ plus is the preference with highest dose and is optimized to produce images with high contrast. IQ Standard is optimized to produce images with low noise and give around 70% of the IQ plus dose. IQ smart produce images with high contrast and radiate the patient with 65% of the IQ plus dose. RDL Plus give 55% of the IQ plus dose and is optimized to give images with high contrast.
RDL Standard is the lowest dose preference with 35% of the dose of IQ plus and is optimized to produce images with low noise [18].
• Two different Dose Reduction Strategies exist in the system. The first one is called Balanced IQ/Dose that reduce the dose up to 25% at 15 fps and 44% at 7.5 fps in fluoroscopic mode. The second is called Maximum Dose Reduction and will reduce the dose up to 50% at 15 fps and 75% at 7.5 fps. All these values are compared with the doses from the 30 fps setting and can only be changed in fluoroscopic mode [18].
2.5 Diagnostic value
Diagnostic value is a subjective measurement and indicates the ability to find an abnormality within an image. The diagnostic value depends on the image quality of the image produced. The image quality depends on the resolution, contrast and the level of noise in the image.
Spatial resolution measures the ability to distinguish the edge between two ob- jects with different intensity and it depends on the geometrical sharpness and the movement of the object. The geometrical sharpness depends on the Source to Image Distance (SID), Object to Image Distance (OID) and focal spot size. Due to the fact that the focal spot is not a point source, the best resolution will be obtained with a small focal spot, a large Focus to Image Distances (FID) and small OID. Keeping the FID and the OID large will also minimize the affect of image distortion, like magnification, in the image. The movement is a problem when imaging the heart and it is solved by using fluoroscopy and recording sequences when imaging the heart [21].
The definition of contrast is the ability to distinguish different objects within the image. When the radiation goes trough the tissue the radiation is absorbed dif- ferently depending on its attenuation coefficient. The thickness and the attenuation coefficient at a certain energy level will decide how much the outgoing intensity is reduced [16]. The Signal to Noise Ratio (SNR) is defined as:
∆S = D2− D1, SN R = ∆S σN
where D1 is the background intensity, D2 is the target intensity and σN is the standard deviation of the noise. The SNR should be as high as possible in order to get a high contrast image. Increasing the radiation delivered to the patient will reduce the effect of random fluctuations in the image called quantum noise and increase contrast for small objects and also increase the resolution. The SNR will be used as a measurement of diagnostic value in this thesis [21].
Method
3.1 Workflow - Vrinnevi Hospital
3.1.1 Pre-operation
Before a patient goes through a coronary angiography, a couple of investigating steps are performed to minimize the risk of complication during the procedure. The first step is to do an Electrocardiography (ECG) and take a blood sample to find out if there are any risks for complication due to the contrast solution injected during the procedure. If the patient has a normal ECG and has not had previous Coronary Artery Bypass Graft (CABG)or heart infarction, no further investigation has to be done. If the patient is suffering from diabetes type two and takes medicine for it, the treatment should be stopped during the time for the coronary angiography due to the risk of kidney failure. Patients with a low Glomerular Filtration Rate (GFR) are asked to drink water before the procedure. GFR is a method to measure the function of the kidneys from a blood sample. Medicine for allergy against the contrast solution is administrated if needed.
A patient that arrives to the hospital is directed to the monitoring department where the nurses ask questions about the patients medical and hygienic conditions.
Pre-medication will be given on the morning before the interventional procedure.
This pre-medication includes medicine for the pain and anticoagulation. The patient is told to put on a hospital shirt, socks and underwear and is then transported in a bed to the catheterization laboratory (cath-lab). A check-list that all of these criteria are fulfilled will be sent with the patient to the cath-lab.
3.1.2 Operation
The goal is to perform all angiography of the coronary arteries through the radial artery. This includes both the elderly and the emergency patients. Before beginning the examination an allen test or an oximetry test is performed to see that the radial artery is refilled correctly.
The procedure starts with injecting local anaesthesia in the skin of the puncture
site. A small hole is cut in the skin with a scalpel and the artery is then punctured with a small needle and a guidewire is introduced. The needle is then removed.
The introducer then slides over the guidewire and a solution of different medicines is injected through the side arm of the introducer and the guidewire is guided into the aortic root.
A catheter is then inserted into the introducer over the guidewire. The insertion of the guidewire, introducer and the canulation of the coronary arteries is monitored using low dose fluoroscopy.
When the operator have entered the coronary artery angiography images can be produced and used for diagnosis. These images are produced using cine mode.
The usage of contrast solution and radiation should be as low as possible and therefore the following projections are used at Vrinnevi Hospital to achieve a good diagnose with minimal dose.
For visualization of the left coronary artery three projection are used (See Figure 3.1). The first one is Caudal 25◦/RAO 15◦ which gives a good visualization over the main left branch and LCX. The second projection is Caudal 35◦/LAO 45◦ this projection is called Spider and gives a good visualization over the bifurcation of the LCX and LAD. The third projection is Cranial 40◦/LAO 10◦ and this projection gives good images representing LAD and the small vessel that originate from it.
Figure 3.1: The three projection used to visualise the left coronary artery.
Left:Caudal 25◦/ RAO 15◦. Middle: Caudal 35◦/LAO 45◦. Right: Cranial 40◦/LAO 10◦.
The visualization of the right coronary artery is performed using two projections (See Figure 3.2). The first projection is LAO 30◦ that gives a good visualisation of the beginning and the middle part of the right coronary artery. The second projec- tion Caudal 10◦/AP gives a good view of the end of the right coronary artery.
Figure 3.2: The two projection used to visualise the right coronary artery. Left:LAO 30◦. Right:Caudal 10◦/AP
PCI
The guidance of the balloon to the lesion is done using fluoroscopy (see Figure 3.3).
When the balloon is deflated and the vessel extended so that blood can flow freely again, two sequences are taken in different directions. The images are taken to investigate if the balloon was inflated enough and if the vessel has its normal size again. The same procedure is done if the stent covered balloon is used.
Figure 3.3: The procedure to perform stenting. The top left image show the stenosis and the top right image show how the stent catheter is progressed. The left bottom image show how the placement of the stent is done and the right bottom show the result after the stent placement. [7]
3.1.3 Post-operation
After the procedure is performed a TR-band (Terumoband) is placed on the in- sertion site to stop the bleeding by putting pressure on the wound. The pressure is created by inflating air into the TR-band. The patient is transported back to the monitoring department in a bed and is allowed to eat and drink immediately
afterwards. The nurses are observant of any chest pain and the patient will be monitored using ECG. The patients normal medication is resumed with additional medication for the pain or other symptoms as a consequence of the procedure.
The pressure in the TR-band will successively be deflated and is taken away after three to four hours. After removing the TR-band, the insertion site is inspected. If the patient is fine and the arm looks normal the patient can go home.
The information about the procedure is written by the operator in the hospital RIS (Radiology Information System) and in SCAAR/Swedheart. The images are stored in PACS. After some time the patient returns to the hospital for a revisit to see if there are any additional problems.
3.2 Literature study
The goal of the literature study is to investigate the factors that affect the dose received by the clinical staff and the patient during a procedure. The factors are separated into three different categories depending on who/what is able to reduce the dose. The categories are separated into technical, operator and patient depen- dency.
3.2.1 Operator dependent
The operator has the ability to control many of the factors that increase the exposure during a PCI procedure. The importance of training, both when it comes to the usage of the x-ray machine and performing a PCI procedure, should therefore be continuously updated. A 60% dose reduction have been measured between a first and a second year cardiologist. The big difference between an experienced doctor and a less experienced doctor is longer fluoroscopic time because of the difficulty of positioning the catheters [22].
There are some things that easily can be changed by the operator that will minimize the dose. The first thing is the frame rate, that should be as low as possible without loosing the clinical value in the images. Pulsed x-ray is used when performing a PCI with a fluoroscopic system, lower frame rate reduces the active time that the x-rays are delivered and lower frame rate gives lower x-ray exposure [18].
The second thing is to minimize the usage of magnification, which is the same as smaller Field Of View (FOV). When using lower FOV the dose increases to compensate for the higher level of noise in the image. By using the largest FOV and then using collimation the additional dose is minimized [23]. Collimation is a good way to reduce the contribution of scattered radiation in the produced image and it will also decrease the scattered radiation delivered to the clinicians. The collimator is situated directly after the tube and limits the area which is radiated, which directly decreases the DAP delivered to the patient [24] [18].
The fourth thing that the operator can change to decrease the dose delivered to the patient is to minimize the distance between the patient and the detector. The distance between the x-ray tube and the patient should be maximized to reduce the dose delivered to the skin and the table should be as high as possible [25] [23].
Using the setting described will also decrease the distortion of magnification in the image [21].
Cine mode is often used when producing diagnostic images. This mode expose the patient up to ten times more than using normal fluoroscopic mode [23]. There- fore the number of frames taken in cine mode should be as few as possible. Step projection of the C-arm increases the thickness of the medium that the radiation needs to pass through (people are often wider than thicker) to reach the detec- tor. The increased dose is because the automatic dose control is activated and will increase the dose because of the reduction of photon reaching the detector [26] [23].
The choice of insertion site also affects the dose. The most used insertion sites for PCI are through the femoral or radial artery. Many studies reports a higher dose exposure to the operator during a radial procedure compared with a femoral.
The reason is the reduced distance to the x-ray tube and longer fluoroscopic time due to the more complicated procedure [22].
The best way to avoid damages from radiation for the clinicians is to increase the distance from the radiation source. The dose delivered reduces inversely pro- portional to the square of the distance [22].
The second best way of reducing the risk of damages is to use radio protective shields and cloths. The radio protective protection is made out of highly absorbing materials that absorbs the x-rays before interacting with the tissue. The radio protective shields are often ceiling-suspended and are placed between the operator and the main source of the scattered radiation which is the patient and detector. A table side shield can be mounted and is used for protection from scattered radiation going downwards from the patient and table towards the operator’s crotch, legs and feet [22].
3.2.2 Patient dependent
The patient dependent factors cannot be modified by the operator but are very important to be aware of. The correlation between Body Mass Index (BMI) and the radiation dose delivered during a PCI is very strong. Articles show that an obese (BM I > 30kg/m2) patient need more than two times the dose of a patient with normal BMI [27]. The dose has to be increased for the x-rays to be able to penetrate the thicker patient and produce diagnostic images [28].
Higher complexity of the procedure increases the radiation dose delivered to the patient during the procedure [22]. PCI procedures that are performed on patients with Chronic Total Occlusion (CTO), bifurcation (stenos in a place where the vessel is divided) or a multi vessel disease need more fluoroscopic time, due to the increased difficulties of catheters placement [22].
Additional equipment for guidance or imaging of the stenosis might be necessary
during the procedure. Intra Vascular Ultrasound (IVUS) and Optical Coherence To- mography (OCT) are used as imaging tools from the inside of the vessel. Fractional Flow Reserve (FFR) and Instantaneous wave-Free Ratio (iFR) are often used to detect if a stenosis should be treated with PCI or not. These techniques needs radiation to guide the catheters used to the right place [28].
3.2.3 Technically dependent
The most efficient method to reduce scattered radiation in the image is to add an anti-scattered grid between the patient and the detector. The grid absorbs the scattered radiation before it reaches the detector. The usage of an anti-scattered grid increases the dose because of the reduction of photons reaching the detector. Using a grid is therefore a trade-off between better image quality and higher dose [18].
A filter can be placed between the x-ray tube and the patient. In the Innova both aluminium and copper filters are used. The filters absorb the soft (low energetic photons) x-rays that only increase the skin dose. The photons that are transmitted through the filters are the ones that contributes to the image [18].
The usage of a digital detector reduces the dose to the patient. The digital detector have a higher Detector Quantum efficiency (DQE) which provides a higher image quality at the same dose [18]. DQE is a measurement method showing how the signal to noise ratio (SNR) will vary from the detection of photons to a produced image. The SNR will increase after the detection in the detector because of quantum noise in the electronics. DQE is a measurement of the accuracy of the image chain.
DQE = (SN Rout/SN Rin)2 [21] The higher DQE permits the operator to use lower dose to obtain the same image [18].
3.3 Pre-study
The x-ray equipment used in the pre-study and the pilot study is a General Electric Innova IGS 520 Medical x-ray angiography equipment with digital detector.
The radiation measurements in this thesis are mainly taken from the data showed in the system monitors. These values are based on the x-ray generators kVp, mAs, spectral filtration, FOV and collimation. From these values the dose rate and the DAP value is calculated and showed by the system [18].
The Pre-study consists of six different trials that investigate how the change of different parameters impact the radiation dose delivered by the system and how the changes impact the image quality.
3.3.1 Basic elements of the pre-study
All trials in the pre-study are performed with the table fixed in Internvetional Reference Point (IRP) positioned 15 cm bellow isocenter (57 cm from the floor).
The Source Image Distance (SID) is fixed at 110 cm. All trials use the same dose protocol called Coronaries, frame rate 7.5 second in fluoroscopic mode, frame rate
15 frames per second in cine and 20 cm FOV (see Figure 3.4). If there are any deviations from this set-up it will be mentioned in the method for that specific trial.
Figure 3.4: Height of detector and table
To simulate patient thickness PolyMethyl MethacrylAte (PMMA) is used. Two types of PMMA phantoms are used in this study where the size of the phantoms are 25x30 cm and 30x30 cm with a thickness of 5 cm (see Figure 3.5). In many trials of the pre-study the thickness of PMMA varies between 5-40 cm.
Figure 3.5: Two types of PMMA phantoms: Right 30x30 cm, Left 25x30 cm
The variation of image quality throughout the trials is analysed using a NRT standard phantom (see Figure 3.6). The image plate consists of different figures with different characteristics (see Figure 3.7). The analyse of image quality is based on images taken from the end of each series. The image quality is evaluated by calculating and comparing the Signal to Noise Ratio (SNR). The SNR is evaluated from two circular objects called Region Of Interest (ROI) on the test phantom. The
ROIs are the same size and are measured to an area of 38, 92 pixel2. The place where the ROI is measured varies between the trials and is shown in Figure 3.7.
Figure 3.6: NRT standard phantom GE medical system
The SNR is calculated using the formula SN R = ∆ ¯σI
B where ¯I is the subtraction between the average intensity of the object and the background ∆ ¯I = ¯IO − ¯IB divided by the standard deviation of the of the signal noise σO [21]. The SNR is used as a measurement for the diagnostic value of the images. All values for the calculation of the SNR from the ROIs are obtained using a program called ImageJ.
The dose rate and SNR is then calculated and plotted using Excel.
Figure 3.7: Left: ROI used in the patient thickness, frame rate and dose protocol trial. Right: ROI used in the OID, FOV and collimation trial.
3.3.2 Patient thickness
The impact of patient thickness is measured using 5-40 cm of PMMA. The trial begins with 5 cm of PMMA that is added on top of the NRT image plate. The image plate and the first layer of PMMA is then radiated using both cine and fluoroscopic
mode. All images and dose rate values are saved before the next layer of PMMA is added. The same procedure is then performed when successively increasing the PMMA thickness to the maximum of 40 cm (see Figure 3.8). The images are used for the calculation of the SNR. The SNR and dose rate will be plotted and used for the evaluation of how increased patient thickness effect the diagnostic value and the dose rate.
Figure 3.8: Variation of PMMA thickness throughout the measurements
3.3.3 Field Of View (FOV)
The Field Of View (FOV) is tested using 5-40 cm of PMMA. The trial begins with FOV 20 cm and 5 cm of PMMA that is added on top of the NRT image plate. The image plate and the PMMA is the imaged using both cine and fluoroscopic mode.
All images and dose rate values are saved before the next layer of PMMA is added.
The same procedure is performed while successively increasing the PMMA thickness to the maximum of 40 cm. The same procedure is then performed with FOV: 17 cm, 15 cm, 12 cm (see Figure 3.9). The images are used for the calculation of the SNR. The SNR and dose rate is plotted and used for evaluation of how decreased FOV effect the diagnostic value and the dose rate.
Figure 3.9: The different FOV parameters
3.3.4 Object to Image Distance (OID)
The impact on dose when changing the OID is tested using different thickness of PMMA (15-40 cm). The trials are performed with the SID changed between 119-86 cm when radiating the PMMA and NRT phantom with x-rays in both fluoroscopic and cine mode.
The measurement starts with the detector in the top position (SID: 119 cm) and 15 cm of PMMA. The OID is then successively decreased according to Table 3.1.
5 cm of PMMA is then added and the same procedure is performed until reaching the maximum of 40 cm. The dose rate and images are recorded for all trials. The images are used for the calculation of the SNR. The SNR and dose rate is plotted and used for the evaluation of how different OID effect the diagnostic value and the dose rate.
PMMA thickness [CM] SID: 86 90 95 100 105 110 115 119
15 86 90 95 100 105 110 115 119
20 90 95 100 105 110 115 119
25 95 100 105 110 115 119
30 100 105 110 115 119
35 105 110 115 119
40 110 115 119
Table 3.1: SID during the trials
3.3.5 Frame rate
The impact on dose when changing the frame rate is tested using 5-40 cm of PMMA.
The trial begins with 5 cm of PMMA and cine mode with frame rate 30 frames per second (fps). Then the different frame rate is successively tested (fluoroscopy:
30fps, Cine: 15 fps, fluoroscopy: 15 fps, fluoroscopy: 7.5 fps, fluoroscopy: 3.75 fps, see figure 3.10). The same procedure is then performed with additional 5 cm PMMA until reaching the maximum of 40 cm PMMA. The dose rate and images are recorded for all trials. The SNR is then calculated using the saved images from the trials. The SNR and dose rate is plotted and used for the evaluation of how different frame rates effect the diagnostic value and the dose rate.
Figure 3.10: The different frame rate settings
3.3.6 Collimation
The size of the x-ray field during different collimation is controlled and measured using the equipment showed in Figure 3.11.
Figure 3.11: The set-up for measuring the size of the X-ray field
The set-up consists of a 3 mm thick copper plate, 25x25 cm cardboard sheet, tape and a Unfors DXR+. The copper plate is taped on the back of the cardboard
plate which is then taped to the detector. The copper plate is used to make the delivered kV of the system higher, making it possible for the Unfors DXR+ to detect a signal. The center of the cardboard plate is marked, together with its 5 cm and 10 cm line. The red line of the Unfors DXR+ is centred on the 10 cm or the 5 cm line depending on the size of the field.
The measurements are performed on both sides (top-bottom, right-left) to min- imize the effect of bad centralization of the cardboard plate (see Figure 3.12). The field size is then calculated by adding the distance in between the measure points (5 cm and 10 cm line) and the values showed on the ruler in the two positions.
Figure 3.12: Measurement pattern
The different field sizes tested is 18 cm, 15 cm, 12 cm, 9 cm and 6 cm. The trial starts with a field size of 18 cm and 5 cm PMMA on top of the NRT phantom. This set-up is radiated using both fluoroscopic and cine mode using the dose protocol called Coronaries. The PMMA is then successively increased with 5 cm in each step to the maximum of 40 cm. The field size is then minimized and the same procedure is performed. The dose rate and images are recorded for all trials. The SNR is then calculated using the saved images from each the trials. The SNR and dose rate is plotted and used for the evaluation of how different field sizes effect the diagnostic value and the dose rate.
3.3.7 Dose protocol
The settings of the dose protocols tested in this study are described in Table 3.2.
The protocols are evaluated using different thickness of PMMA (5-40 cm). The trial start with 5 cm of PMMA on top the NRT phantom. This set-up is radiated using both fluoroscopic and cine using the dose protocol called Coronaries+. The PMMA is then successively increased with additional 5 cm in each step until the maximum of 40 cm is reached. The same procedure is then performed with the dose protocols Coronaries, Test 1, Test 2, Test 3, Test 4. The dose rate and images are recorded for all trials. The SNR is then calculated using the saved images from the trials.
The SNR and dose rate is plotted and used for the evaluation of how the different dose protocols effect the diagnostic value and the dose rate.
Fluoroscopic mode Baseline (Coronaries+) Baseline (Coronaries) Test 1 Test 2 Test 3 Test 4 Strategy Balanced IQ Balanced IQ Max dosereduction Balanced IQ Balanced IQ Balanced IQ
Frame/sec 7,5 7,5 7.5 3.75 7.5 7,5
Detail Low Low Low Low Low Low
Preferences RDL Plus RDL Plus RDL Plus RDL Plus RDL Standard RDL Plus
Recording mode
Frame/sec 15 15 15 15 15 15
Detail Low Low Low Low Low Low
Preferences IQ standard RDL Plus RDL Standard RDL Standard RDL Standard RDL Standard
Table 3.2: Protocol testing schedule, the bold parameters in the table is the ones that have been adjusted.
3.4 Pilot Study
After evaluating present work flow and dose settings a discussion with the clinical staff and the responsible physicist gave the idea to change the dose protocol and see how this would affect the dose delivered to the patients.
A first step was to evaluate if the protocol produced in the pre-study was useful in a real clinical environment. The doctor systematically tried out the protocols one by one. The evaluation resulted in an acceptance of implementing a new protocol called Test 4. The change compared with the original dose protocol (coronaries) and the Test 4 protocol is that the AutoEx preferences setting in cine mode is set to RDL standard instead of RDL plus.
3.4.1 Statistics
A pilot study is a small study with the goal to investigate the feasibility of an approach that will be used in a larger study called a pivotal study. Hypothesis testing is not very effective when it comes to pilot studies due to the small sample size and the results/conclusions should be considered with caution [29].
Performing a hypothesis testing in a pilot study with small sample size can be done using statistical methods like Mann–Whitney U test for continuous data and Fischer’s exact test for categorical data [29].
Mann-whitney U test is a non-parametric method to test data from two in- dependent groups of continuous data. The Mann-whitney U test requires all the observations to be ranked as if they were from a single sample. The lowest value gets rank number 1 and the second lowest rank number 2 and so on. Each data point is given a score for every data point from the other group that has a higher rank value. The score from each group is summarized and the lowest of the two sums=U. Then z is calculated from the formula [30]:
z = 12(U − nAnB) pnAnB(nA+ nB+ 1
Where nA is the amount of data points in group A and nB amount of data points in group B . The z value is then compared with a z-table and the p-value can be found. From this p-value the null hypothesis is rejected at a certain pre-decided level [30].
Table 3.3: Example of a test matrix for a Fischer’s exact test.
Control group Test group Sum row
Male a b a+b
Female c d c+d
Sum column a+c b+d N=a+b+c+d
Table 3.4: Example of a test matrix for a Fischer’s exact test.
Fischer’s exact test is used for small samples of categorical data. The method is performed by evaluating a 2x2 matrix of observed data, see an example of the matrix in Table 3.3. The null hypothesis is that there are no associations between the rows and the column variables [30].
A p-value that show if the null hypothesis should be kept or rejected is then calculated from the formula:
p = (a + b)!(a + c)!(b + d)!(c + d)!
N !a!b!c!d!
In this thesis Matlab is used for calculation of the p-value for both statistical models.
3.4.2 Radiation dose measurements
The radiation dose delivered by the system during the procedure is mainly repre- sented by the DAP-value but the fluoroscopic time is also considered. The fluoro- copy is used during the guidance during the procedure but it does not include the energy delivered to the patient. The DAP represent the dose delivered to the pa- tient. The DAP value show the amount of dose delivered to the patient multiplied with the irradiated field. The DAP value is an indication for the stochastic effect of the radiation [31].
3.4.3 Baseline comparison
The retrospective data is gathered in the period 2015.03.04-2015.10.08 and the information include both angiography and PCI. The data is taken from procedures done by the same doctor and all procedures are performed on the same machine.
The retrospective data will be compared with the data from patients in the period 2015.11.02-2015.12.02 using dose protocol Test 4.
The data before the change and the data received after the change of the protocol are presented using percentage for categorical data. Continuous data are presented using median with its 1st,3rd quartile and maximum/minimum value. This presen- tation of the data is chosen because it gives a good overview of the distribution of
the datasets and is therefore easy to understand. All calculation of data is done in Excel.
The analysed parameters are:
• Age
• Gender
• BMI
• Prior CABG
• Prior PCI
• Access point
• Number of stenosis
• Number of stents
• Usage of additional techniques
• CTO
• Contrast
• Fluoroscopic time
• DAP
Results
4.1 Pre-study
4.1.1 Patient thickness
Dose rate
Figure 4.1 shows how the dose rate varies with the PMMA thickness that is used as a simulation of patient thickness in this thesis. The left graph shows the dose rate in fluoroscopic mode and the right one is in cine mode. In the end of both series, the x-ray machine reaches its output maximum for that protocol and the derivative of the curve approaches zero.
Figure 4.1: Left: Dose rate in fluoroscopic mode. Right: Dose rate in cine mode
Image quality analysis
Figure 4.2 shows how the SNR varies with the PMMA thickness. The left graph shows the SNR in fluoroscopic mode and the right one is in cine mode. Both graphs show decreased SNR with higher PMMA thickness.
