UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Source strength verification and quality assurance of sterile, pre-loaded iodine-125 seed trains used… Afsarigolshan, Maryam 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2014_spring_afsarigolshan_maryam.pdf [ 2.56MB ]
Metadata
JSON: 24-1.0167373.json
JSON-LD: 24-1.0167373-ld.json
RDF/XML (Pretty): 24-1.0167373-rdf.xml
RDF/JSON: 24-1.0167373-rdf.json
Turtle: 24-1.0167373-turtle.txt
N-Triples: 24-1.0167373-rdf-ntriples.txt
Original Record: 24-1.0167373-source.json
Full Text
24-1.0167373-fulltext.txt
Citation
24-1.0167373.ris

Full Text

SOURCE STRENGTH VERIFICATION AND QUALITY ASSURANCE OF STERILE, PRE-LOADED IODINE-125 SEED TRAINS USED FOR PROSTATE BRACHYTHERAPY  by  MARYAM AFSARIGOLSHAN   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE  in  THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Physics)   THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)   April 2014   © MARYAM AFSARIGOLSHAN, 2014  iiAbstract  Like all brachytherapy sources, radioactive “seeds” that are supplied to the clinic in pre-loaded needles require quality assurance measures that include, if possible, verification of the individual strengths of all the seeds.  Currently, there is no commercially available apparatus or established technique that allows such measurements to be made on seeds already loaded into needles. Source strength verification by the physicist receiving the order is generally limited to assaying seeds from a separate sample belonging to the same batch that was used to load the needles. However, this method does not assay the seeds that will actually be implanted into the patient, and there is potential for errors to occur between the time the needle loading service assays the seeds and the time that the loaded needles and verification seeds are packaged and sent to the clinic.  We investigated a method to verify the positions and strengths of brachytherapy seeds loaded into implant needles. The system used consisted of a flat panel image detector, an autoclavable apparatus (“jig”) to hold the needles in a fixed geometry relative to the detector, and software to analyze the detected signals. In addition to defining the geometry of the detection system, the jig serves to maintain a sterile field around the needles as they are being assayed. The goal will be for this system to provide relative dosimetric readings that can be compared to readings from needles loaded with seeds of known strength, as established by measurement in a well chamber. Results showed a good correlation between detector reading and radioactive source strength of up to 6 individual seeds loaded into needles as trains with 1 cm or greater spacing between seeds. In addition, the detector has a sufficiently large image capture area to allow several needles (10-15) to be analyzed simultaneously. Good correlation between detector reading and exposure time, as well as between detector reading and source strength, demonstrated the linear characteristic of the detector.  This study showed that a flat panel detector dosimetry system allowed simultaneous characterization in a sterile environment of 100% of seeds used in a brachytherapy treatment procedure.    iiiPreface  The original concept, including use of an imaging device and collimation system to extend the distance between source and detector belongs to Dr. Ingrid Spadinger. The first jig apparatus shown in Figure.18 was designed by Dr. Nick Chng. My original contributions to this work were design of the second jig apparatus, all the experiments and experimental designs, evaluation of the image detector and developing the analysis algorithms.                           ivTable of Contents  Abstract .................................................................................................................................... ii Preface ..................................................................................................................................... iii Table of Contents ................................................................................................................... iv List of Tables .......................................................................................................................... vi List of Figures ........................................................................................................................ vii Acknowledgements ................................................................................................................ ix Dedication ................................................................................................................................ x Chapter  1  Introduction ..................................................................................................... 1 1.1  Clinical Background ............................................................................................................. 1 1.1.1  Incidence of Prostate Cancer ............................................................................................ 1 1.1.2  Prostate Anatomy ............................................................................................................. 1 1.1.3  Diagnosis and Staging ...................................................................................................... 1 1.1.4  Treatment .......................................................................................................................... 3 1.2  Overview of Brachytherapy Physics ..................................................................................... 4 1.2.1  Radioactive decay ............................................................................................................. 6 1.2.2  Brachytherapy Sources ..................................................................................................... 7 1.2.3  Brachytherapy Dosimetry ................................................................................................. 8 1.2.3.1  Interactions of Photons with Matter ......................................................................... 8 1.2.3.1.1  Coherent Scattering ............................................................................................. 9 1.2.3.1.2  Photoelectric Effect ............................................................................................. 9 1.2.3.1.3  Compton Effect ................................................................................................... 9 1.2.3.2  Air Kerma Strength and Absorbed Dose ............................................................... 12 1.2.3.3  Dosimetry Formalism ............................................................................................ 13 1.2.3.4  Activity .................................................................................................................. 15 1.3  Treatment Planning for Permanent Implant Prostate Brachytherapy .................................. 16 1.3.1  Target Volume Definition ............................................................................................... 16 1.3.2  Planning Approaches ...................................................................................................... 17 1.3.3  Needle and Seed Placement Approaches ........................................................................ 19  v1.3.3.1  Uniform Loading ................................................................................................... 20 1.3.3.2  Modified Uniform Loading .................................................................................... 23 1.3.3.3  Peripheral Loading ................................................................................................. 23 1.4  Brachytherapy Source Train Assembly and Needle Preparation ........................................ 27 1.5  Calibration of Brachytherapy Sources ................................................................................ 29 1.5.1  Calibration Methods ....................................................................................................... 29 1.5.2  Quality Assurance Standards for Source Strength Verification ..................................... 30 1.5.3  Statistical Considerations ................................................................................................ 31 1.5.4  Calibration of Sources in Sterile Needles, Cartridges, or Strands .................................. 32 1.5.5  Third-party Brachytherapy Source Calibrations ............................................................. 33 1.6  Purpose of this Thesis ......................................................................................................... 35 Chapter  2  Materials and Methods ................................................................................. 36 2.1  I125 Sources .......................................................................................................................... 36 2.2  Radiation Detector .............................................................................................................. 36 2.2.1  Flat Panel Detector Technology ..................................................................................... 37 2.2.2  Choice of Flat Panel Detector ......................................................................................... 39 2.2.3  Imaging System Set-up ................................................................................................... 42 2.3  Imaging Geometry .............................................................................................................. 44 2.3.1  Geometrical Factor Uncertainty ..................................................................................... 45 2.3.2  Additional Sources of Uncertainty ................................................................................. 49 2.3.3  Design of the Needle Holder .......................................................................................... 49 2.4  Image Data Acquisition ...................................................................................................... 52 2.5  Seed and Needle Configurations ......................................................................................... 53 2.6  Seed Parameters and Verification ....................................................................................... 53 2.7  Characterization of the System Response ........................................................................... 54 2.8  Image Analysis .................................................................................................................... 56 Chapter  3  Results ............................................................................................................ 58 3.1  Results ................................................................................................................................. 58 Chapter  4  Conclusion ..................................................................................................... 64 References .............................................................................................................................. 66   viList of Tables   Table 1: American Brachytherapy Society (ABS) Nomenclature for Different Types of Planning Used in Permanent Prostate Brachytherapy ............................................................ 17 Table 2: Quantities of brachytherapy sources to be assayed by the end-user physicist ......... 34 Table 3: Technical data 1512 .................................................................................................. 40 Table 4: 18 and 20 gauge needles information ....................................................................... 47 Table 5: Seed position uncertainty relative to detector surface .............................................. 48 Table 6: Attenuation effect due to variable wall thickness of the needle ............................... 49 Table 7: Temporal test result .................................................................................................. 60 Table 8: Comparison of the seeds activity calculated by two different methods ................... 62   viiList of Figures  Figure 1: Relative contributions of different photon interactions in water and stainless steel9................................................................................................................................................. 11 Figure 2: Coordinate system used for brachytherapy dosimetry calculation .......................... 14 Figure 3: Stepper system and template ................................................................................... 18 Figure 4: DVH report of uniform loading ............................................................................... 21 Figure 5: Uniform loading treatment plan .............................................................................. 22 Figure 6: Needle loading report .............................................................................................. 24 Figure 7: Seed position report ................................................................................................. 25 Figure 8: Modified loading treatment plan ............................................................................. 26 Figure 9: example of seed configurations ............................................................................... 27 Figure 10: Example of a stranded seed train prior to insertion into the needle ...................... 28 Figure 11: Schematic diagram of a well type chamber ........................................................... 30 Figure 12: Schematic diagram of I125 seeds, Model 6711 ...................................................... 36 Figure 13:  Schematic diagram of CMOS detector ................................................................. 38 Figure 14: Enclosure Dimensions (1512) ............................................................................... 41 Figure 15: example of needle autoradiograph ........................................................................ 45 Figure 16: Discrepancy of geometrical factor for 0.1mm distance uncertainty ..................... 46 Figure 17: Seed position uncertainty relative to detector surface ........................................... 47 Figure 18: Wedged jig design ................................................................................................. 50 Figure 19: Stainless steel collimator used in the experiments ................................................ 51 Figure 20: Image taken with old jig (Right), new jig (Left) ................................................... 51 Figure 21: Image of 4 seeds with apparent activity of 0.292 mCi .......................................... 53 Figure 22: Orientation test set up ............................................................................................ 54 Figure 23: Image analysis ....................................................................................................... 56 Figure 24: Detector reading in two different directions of a needle ....................................... 58 Figure 25: Shifting the seeds along the needle ....................................................................... 59 Figure 26: Rotating the needle ................................................................................................ 59 Figure 27: Measuring a needle with 4 seeds 10 times ............................................................ 60 Figure 28:  Changes in detector reading resulting from different exposure times .................. 61  viiiFigure 29: calibration graph for three different regions of the detector ................................. 61 Figure 30: Activity measured by well ion chamber vs. activity estimated by calibration graph................................................................................................................................................. 63                              ixAcknowledgements  I would like to thank my supervisor, Dr. Ingrid Spadinger, for all her support, close supervision and helpful answers to my many questions and specially her insightful comments on my work without which this thesis would not have been possible. I would also like to thank Dr. Nick Chng who was always willing to help and give his best suggestions. I would like to extend my sincere thanks to all of the Medical Physics staff at BC Cancer Agency- Vancouver center for their guidance and support. I would like to acknowledge BC Cancer Agency prostate brachytherapy program for their financial support.  xDedication   To my beloved Parents  To Lovely Tamara and Taimaz  To my love Afshin  1Chapter  1  Introduction  1.1 Clinical Background  1.1.1 Incidence of Prostate Cancer  Prostate cancer is the most common cancer diagnosed in men, with 23,600 cases in 2013. It is the third most common cause of cancer death in males with approximately 4000 deaths. Prostate cancer will be diagnosed most frequently in males in their 60s, but more prostate cancer deaths will occur in males 80 years and older which reflects the effect of screening men for prostate cancer in their 60s and the long natural history of the disease.1  The three main treatments for the disease are prostatectomy (surgery), external beam radiotherapy and implantation of radioactive sources (brachytherapy).   1.1.2 Prostate Anatomy  The prostate is part of the male reproductive system. It is about the same size and shape as a chestnut. It is located below the bladder and in front of the rectum, and surrounds the superior portion of the urethra. It secretes an alkaline fluid that makes up about 13% to 33% of semen. This fluid plays a role in activating the sperm cells to swim.  1.1.3 Diagnosis and Staging  There might be no symptoms in the early stages of the disease. As men age, the prostate can enlarge due to several reasons. In the case of these symptoms patients should be checked by a physician: problem in starting or stopping urination, painful or frequent urination, loss of urinary control and blood in urine. These symptoms might be the result of the benign swelling of the gland over the course of normal aging, but can also be caused by prostate cancer. Detecting the cancer at an early stage increases the chance of cure. Currently, common detection methods are screening the prostate specific antigen (PSA) levels and digital rectal examination (DRE).    2The PSA test measures the amount of prostate specific antigen in the blood. The results usually reported as nanograms of PSA per milliliter (ng/mL) of blood. Prostate specific antigen is a protein secreted by the epithelial cells of the prostate gland that liquefies semen. Typically, cancerous prostate glands release more PSA into the blood circulation than healthy prostate glands. In addition to prostate cancer some benign conditions can cause an elevation in PSA level, like prostatitis (inflammation of the prostate) and benign prostatic hyperplasia (BPH) (enlargement of the prostate). Normal PSA ranges are age-specific and will increase with age. DRE allows the doctor to detect prostate abnormalities by feel. If there is a high PSA level or a lump is felt, confirmation of cancer can be obtained by the pathologic examination of prostate tissue. The decision to proceed to prostate biopsy is based primarily on PSA and DRE results but also takes into account multiple factors like patient age, PSA density, family history, ethnicity and co morbidities. Transrectal, ultrasound guided prostate biopsy is the most common method, where 8 to 12 core tissue biopsies are removed,  usually under local anesthesia2.  If cancer is found, a specialist will determine the size, stage and the grade of the tumor, which will help in determining the type of therapy recommended for treating the cancer. The Gleason Grading system is used to help evaluate the prognosis of men with prostate cancer and is based on architectural growth pattern of prostatic carcinoma. It is the dominant method around the world and was developed by Dr. Donald F Gleason, a pathologist in Minnesota. Five basic grade patterns are used to generate a histologic score from 1 to 5, with 5 having the worst prognosis. The score is typically reported by adding the grade of the two most dominant patterns in the biopsy samples.3 High Gleason score means the cancer tissue is very different from normal and the tumor is more likely to spread. Staging of prostate cancer is a process that is used to find out if the cancer has spread within the prostate or to other parts of the body. The stage is based on the prostate biopsy results (including the Gleason score), the PSA level, and any other exams or tests that were done to find out how far the cancer has spread. The most common staging system is promulgated by the American Joint Committee on Cancer, and is known as the TNM system. It is based on the following information: the size of the primary tumor (T category), the extent of involved lymph nodes (N category) and the absence or presence of distant   3metastasis (M category). It also takes into account the PSA level at the time of diagnosis and the Gleason score. All this information is combined, along with Gleason score and prostate specific antigen, in a process called stage grouping. The overall stage is expressed in Roman numerals from I to IV. In stage I, cancer is found in the prostate only. In stage II, the tumor has grown inside the prostate but hasn’t extended beyond it. In stage III, cancer has spread beyond the outer layer of the prostate and may have spread to the seminal vesicle. In stage IV, cancer has metastasized to other tissues.   1.1.4 Treatment  There are several types of treatment used for patients with prostate cancer. Active surveillance is recommended for patients with small or low grade cancer. It involves frequent monitoring of the tumor and regular DREs and PSA tests. No treatment is given to a patient unless there are changes in test results. Prostatectomy is the surgical removal of all or part of the prostate gland. Retropubic prostatectomy is done through the abdominal wall and nearby lymph nodes may be removed at the same time. In perineal prostatectomy, the prostate surgery is done through an incision made in the perineum, but nearby lymph nodes may be removed through a separate incision in the abdomen.  Radiation therapy uses high energy x-rays to kill cancer cells or stop them from growing. There are two types of radiation therapy: external beam radiotherapy and brachytherapy (internal radiotherapy). Hormone therapy is a treatment based on the reduction of naturally present androgen hormones, which stimulate prostate cancer cells to grow. This therapy is widely used to treat men with clinically localized prostate cancer, biochemical recurrence after radical prostatectomy, locally advanced disease or lymph node metastases.4  Chemotherapy is the use of cytotoxic drugs to destroy cancer cells. It can destroy cancer cells that have metastasized, or spread from the prostate gland to other parts of the body.  In contrast to surgery and radiation therapy that remove, destroy or damage cancer cells in a specific area, chemotherapy works throughout the body via the bloodstream and kills any rapidly growing cells, both cancerous and non-cancerous. Chemotherapy is   4sometimes used if prostate cancer has spread outside the prostate gland and hormone therapy isn't working. It is not a standard treatment for early prostate cancer, but some studies are looking to see if it could be helpful if given for a short time after surgery.  1.2 Overview of Brachytherapy Physics  Brachytherapy is a method of treatment in which sealed (encapsulated) radioactive sources are used to deliver radiation at a short distance. With this mode of therapy, a high radiation dose can be delivered locally to the tumor with rapid dose fall-off in the surrounding normal tissue. In terms of placement of the radioactive sources, brachytherapy may be divided into 3 categories: interstitial brachytherapy, where the sources are implanted in the target tissue directly, intracavitary brachytherapy, where the sources are placed in body cavities at distances on the order of a few millimeters or centimeters from the target tissue, and surface brachytherapy where sources are placed externally on structures such as the eye and skin. Different types of brachytherapy delivery systems have been developed in order to reduce radiation exposure of personnel. In many procedures, the hazards associated with direct “hot” loading have been reduced by the adoption of alternative techniques like manual afterloading and remote-controlled afterloading. In afterloading techniques, hollow needles, catheters or other types of applicators are first inserted into the target volume. Because no radioactive material is present, clinicians can take their time optimizing the position of the applicators without incurring excessive radiation exposure. After insertion is completed, the applicator positions are confirmed using a diagnostic imaging modality such as plane film radiographs, CT, MRI, or ultrasound (CT being the most common), and the treatment plan specifying the distribution of radioactivity is finalized on the basis of this information. The radioactive material is then introduced into the applicators according to the treatment plan. In the case of manual afterloading, radioactive source trains are introduced by hand. Because this process can be done quickly, radiation exposure is minimized. However, such exposure can be virtually eliminated by the use of remote-controlled afterloading, where the applicators are connected to an afterloader machine through connecting guide tubes and   5radioactive source delivery into the pre-specified positions is controlled by a computer program. There are two categories of brachytherapy in terms of duration of implant: permanent and temporary. In a permanent implant, the radioactive sources are permanently implanted into the tumor, the patient is released from the hospital with radioactive material in situ and radioactivity is allowed to decay within the patient. Permanent implants are performed with relatively short half-life radioisotopes like I125 (59.4 days), Pd103 (17 days), or Au198 (2.7 days). Over a period of weeks or months, the level of radiation emitted by the sources will decline to almost zero. Although afterloading can be used to introduce source trains into needles that have been pre-implanted into the tissue, few afterloading systems designed for this purpose are commercially available5. In addition, for permanent implants, the needles must be withdrawn whilst leaving the sources behind. While there is interest in automating this final step 6, such systems are currently only investigational.  Consequently, needles are typically withdrawn manually, and only a few clinics use the afterloading system for needle insertion. More commonly, needles are loaded with the radioactive source trains prior to implantation and these “hot” needles are then inserted into the tissue.  Radiation exposure to the practitioner during such procedures is minimized through the use of low energy sources such as I125, Pd103, and Cs131, which were first introduced in the 1960’s for this purpose.7, 8  In a temporary implant, the radioactive material is temporarily introduced into or close to the tumor and is removed once the desired radiation dose has been delivered. The specific treatment duration will depend on many different factors, including the required rate of dose delivery and the type, size and location of the cancer. Most temporary procedures are delivered using afterloading techniques. Temporary brachytherapy can be delivered at different dose rates, which have been divided into low, medium, and high dose rate categories by the International Commission of Radiation Units and Measurements (ICRU) Report No.38: 1. Low Dose Rate (LDR): 0.4 to 2.0 Gy per Hour  This is the traditional dose rate for manually afterloaded brachytherapy, and also includes permanent implants. Many of the advantages of brachytherapy are attributed to the radiobiology of continuous LDR irradiation.   62. Medium Dose Rate (MDR): 2 to 12 Gy per Hour  This dose rate, also called ‘intermediate dose rate’, is rarely used. 3. High Dose Rate (HDR): More than 12 Gy per Hour  HDR brachytherapy utilizes a very high activity (5-10 Ci) source. Ir192, which can be produced with very high specific activity, is the most commonly used radioisotope for this purpose. Treatment can only be delivered by remote-control afterloading techniques and the usual dose rate is about 100 to 300 Gy per hour, allowing treatments to be delivered in only a few minutes with a single source stepped through a number of treatment positions.  Although the higher dose rate often requires the treatment to be delivered in multiple fractions due to radiobiological factors, many treatment sites can be treated on an outpatient basis.  1.2.1 Radioactive decay  Radioactive isotopes are nuclides that have unstable nuclei that decay, emitting alpha, beta, and sometimes gamma rays. Such isotopes eventually reach stability in the form of non-radioactive isotopes of other chemical elements. There are five different types of radioactive decay: Alpha decay results from an excess of mass. In this type of decay, alpha particles (consisting of two protons and two neutrons) are emitted from the nucleus. An example is the decay of 238U.  β + or positron decay results from an excess of protons. In this type of decay, a positively charged beta particle and a neutrino are emitted from the nucleus. The atomic number decreases by one and the neutron number is increased by one. An example is the decay of radioactive 18F to 18O. β- or negatron decay results from an excess of neutrons. In this type of decay, a negatively charged beta particle and a neutrino are emitted from the nucleus. The atomic number increases by one and the neutron number is reduced by one. An example is the decay of radioactive 14C to stable 14N. Electron capture also results from an excess of protons. In this type of decay, an electron is spontaneously incorporated into the nucleus and a neutrino is emitted from the nucleus. The atomic number decreases by one and the neutron number increases by one. An example is the   7decay of 125I to 125Te. Characteristic photons are emitted due to the vacancy created in one of the electron orbitals. Internal conversion is a radioactive decay process where an excited nucleus interacts electromagnetically with an electron in one of the lower atomic orbitals, causing the electron to be emitted (ejected) from the atom. Thus, in an internal conversion process, a high-energy electron is emitted from the radioactive atom, but not from a nucleon in the nucleus. Instead, the electron is ejected as a result of an interaction between the entire nucleus and an outside electron that interacts with it. Gamma ray emission may follow any of these modes of decay, but does not occur for all radioisotopes. Photons in the form of gamma rays and characteristic x-rays are the emissions of interest for most brachytherapy applications.    1.2.2 Brachytherapy Sources  In the early years of brachytherapy, the naturally occurring radionuclides, radium and radon were used exclusively. After about 50 years, new radionuclides entered the brachytherapy field through production of artificial radionuclides in a nuclear reactor. There is an extensive menu of radionuclides emitting gammas, betas, or neutrons with a wide range of energies and half-lives, used in medicine. For brachytherapy applications, they are encapsulated in sealed, biocompatible capsules made of materials such as titanium and stainless steel. Photon-emitting radionuclides used in brachytherapy include Ra226, Cs137, Ir192, Au198, Cs131, I125 and Pd103. Among these I125, Pd103 and Cs131 are used for permanent prostate brachytherapy, with I125 being the most commonly used. I125 has a half-life of 59.4 days, decaying exclusively by the electron capture process. Characteristic x-rays in the range of 27 to 32 keV are emitted as a consequence of the electron capture and internal conversion processes. The HVL thickness for the photons emitted by encapsulated sources containing this radionuclide is about 0.025 mm of lead, making it relatively easy to shield personnel from its emissions. I125 seeds are used principally for permanent implants, but are also suitable for temporary procedures.  Several manufacturers supply I125 seeds for brachytherapy, and each utilizes a slightly different   8source design.  Most designs are cylindrical and have outer dimensions of 0.8 mm diameter and 4.5-5.0 mm length.  They are often referred to as brachytherapy “seeds”. Pd103 has a half-life of 17 days. It decays via the electron capture process with the emission of characteristic x-rays in the energy range of 20-23 keV, and Auger electrons. The weighted mean photon energy is 20.7 keV, and the HVL of photons emitted from encapsulated sources of Pd103 is about 0.004mm of lead. Cs131 is also an x-ray emitter with, the most prominent peaks in the 29 keV to 34 keV region. It decays by electron capture to Xe131. It has a half-life of 9.7 days, which is considerably shorter than that of the other two radionuclides commonly used in prostate brachytherapy. Cs131 is used in a minority of prostate brachytherapy procedures. Pd103 and Cs131 seeds have the same outer dimensions as I125 seeds.  1.2.3 Brachytherapy Dosimetry  1.2.3.1 Interactions of Photons with Matter  When an x-ray or γ ray photon passes through a medium, it has a certain probability of interacting. This probability depends on the energy of the photon and the physical properties of the medium, and is characterized by the linear attenuation coefficient, μ. Attenuation of a non-divergent, monoenergetic photon beam is described by the exponential equation: N(x) = Noe-μx Where No is the number of photons in the beam before it has entered the attenuating material (i.e. N(x) at x=0) When a photon interacts with an atom, it is either completely absorbed or scattered away from its original direction, with or without energy transfer. Energy transfer interactions cause ionization (ejection of electrons) from the atoms of the absorbing medium. The resulting high-speed electrons travel through the medium, transferring their kinetic energy to surrounding atoms through multiple ionization and excitation events along their paths. If the absorbing medium consists of body tissues, sufficient energy may be deposited within the   9cellular DNA to destroy the reproductive capacity of the affected cell, which is the basis of the therapeutic effect of radiation treatment. Attenuation of a photon beam in the 0-50 keV energy range relevant to this thesis is caused by three major types of interactions with the absorbing material. These processes are coherent scattering, the photoelectric effect, and the Compton effect.   1.2.3.1.1 Coherent Scattering  Coherent scattering, also known as classical scattering or Rayleigh scattering, occurs when a photon briefly excites an atomic electron cloud.  The energy involved in this process is ultimately transferred back to the photon, which ends up being scattered without energy loss to the medium. Scattering angles are typically small. Coherent scattering is more probable in high atomic number materials and with photons of low energy.  1.2.3.1.2 Photoelectric Effect  The photoelectric effect is a phenomenon in which a photon of sufficiently high energy interacts with and ejects one of the orbital electrons from an atom. The photon is totally absorbed by an inner shell electron and this electron (called the photoelectron) is ejected. The energy of the emitted photoelectron equals the difference between the incident photon energy and the electron binding energy. Thus, the photon must have sufficient energy to overcome the electron binding energy.  Outer-shell electrons then fill the inner-shell electron vacancy to stabilize the atom, and the excess energy is emitted as characteristic photon radiation or Auger electrons.   1.2.3.1.3 Compton Effect  In the Compton interaction, incident photons interact with loosely bound valence electrons. It results in a scattered photon and an ejected electron. The scattered photon has   10less energy than that of the incident photon and travels in a new direction. The scattered electron carries the energy lost by the incident photon. The probability of a Compton interaction is proportional to the number of outer-shell electrons available in the medium (electron density).   Graph below shows the relative contributions of the different photon interactions in our energy range of interest in water and stainless steel.   11 Figure 1: Relative contributions of different photon interactions in water and stainless steel9       121.2.3.2 Air Kerma Strength and Absorbed Dose  Photon-emitting brachytherapy sources are specified in terms of air-kerma strength, denoted by Sk, in North America. Sk is defined as the product of air-kerma rate, Kδ(d), in vacuo and due to photons of energy greater than δ, at distance d, and the square of this distance, d2.  ܵ௞ ൌ ܭሶఋ,௔௜௥ሺ݀ሻ. ݀ଶ  The point of Kδ,air specification is located on the transverse-plane of the source (the plane normal to the longitudinal-axis of the cylindrical source, which bisects the radioactivity distribution). The unit of air kerma strength is 1 μGy.m2.h-1, and is often denoted in the literature by the symbol ‘U’ where 1 U = 1 cGy.cm2.h-1= μGy.m2.h-1.   The quantity Sk describes source strength in terms of kerma, Kx, which is the total kinetic energy transferred to charged particles by photon interactions with atoms per unit mass of material x. Although kerma can be specified in any medium x, usually air with (x=air), is assumed for radiation metrology and replaces the now obsolete quantity exposure, X. Absorbed dose to air, Dair, and Sk are closely related:  ܦ௔௜௥ ൌ ܺ. ൬ܹ݁൰ ൌ ܵ௞. ሺ1 െ ݃ሻ ൎ ܵ௞  where (W/e) is the average energy imparted to air per ion pair created and is a constant, independent of photon energy: (W/e) = 33.97 eV/ion pair = 33.97 J/C. The factor g is the fraction of kinetic energy transferred to the medium that is converted back to radiant energy (photons) by the Bremsstrahlung process. The quantity Kx. (1-g) is often called ‘collisional kerma,’ and refers to that component of the transferred energy which is ultimately absorbed by the medium via inelastic secondary charged particle collisions. Since g ≤ 0.001 at brachytherapy energies in tissue-like media, this radiative correction is usually ignored. The equation above assumes secondary charged particle equilibrium (CPE) 10. CPE is realized when kerma remains relatively constant over the secondary electron range,   13assuring that the rates of energy absorption and energy transfer are approximately equal, and that kerma closely approximates absorbed dose. Virtually all brachytherapy dose calculation algorithms and dosimetric analyses assume that CPE exists, and that D ≈ K everywhere. Although it is generally valid, CPE can be expected to break down in the presence of steep dose gradients very close to the source11, near metal-tissue interfaces12, and within the active elements of thin, bounded detectors13.  1.2.3.3 Dosimetry Formalism  The TG-43 protocol 14-16 aims to define a formalism based on dose to water at clinically relevant distances (i.e. on the order of 1 cm). However, it recognizes that practical measurement and verification of source strength is most easily carried out in air at larger distances (~1 m), and therefore retains the use of air kerma for source strength specification. Effects of attenuation, scatter, and absorption by source encapsulation are accounted for by a specific set of parameters, as outlined below. The TG-43 parameter values are determined individually for each type of brachytherapy source model, usually through a combination of measurements and Monte Carlo calculations, and then published as a standard for clinical use.  ܦሶ ሺݎ, ߠሻ ൌ ܵ௞  ܩ௅ሺݎ, ߠሻܩ௅ሺݎ௢, ߠ௢ሻ ݃௅ሺݎሻܨሺݎ, ߠሻ   ሺ݃݁݊݁݉ݎ݋݂ ݈ܽݎሻ ܦሶ ሺݎ, ߠሻ ൌ ܵ௞  ܩ௣ሺݎ, ߠሻܩ௣ሺݎ௢, ߠ௢ሻ ݃௣ሺݎሻሶ׎ഥ௔௡   ሺܽ݊݋݅ݐܽ݉݅ݔ݋ݎ݌݌ሻ  Where r denotes the distance (in centimeters) from the center of the a cylindrical source to the point of interest, ro denotes the reference distance, which is specified to be 1 cm in this protocol, and θ denotes the polar angle specifying the point of interest, P(r,θ), relative to the source longitudinal-axis. The reference angle, θo, defines the source transverse plane, and is specified to be 90o or π/2 radians (Figure below).    14  Figure 2: Coordinate system used for brachytherapy dosimetry calculation   Λ is the dose rate per unit air kerma strength at distance ro = 1 cm along the transverse axis of the source. The reference distance, ro, is comparable to distances of therapeutic interest. Values of Λ are specific to the model (both type of radionuclide and physical source design) of brachytherapy source being considered. G(r, θ) accounts for dose variations due to inverse square law effects, with some consideration of the spatial distribution of activity within the source. It does not account for any photon absorption or scatter in the source structure. The distribution of radioactivity is approximated as a point source or a line source: ܩሺݎ, ߠሻ ൌ1ݎଶ  ሺ݁ܿݎݑ݋ݏ ݐ݊݅݋݌ሻ ܩሺݎ, ߠሻ ൌߚߠ݊݅ݏ ݎ ܮሺߠ ് 0ሻ ݎ݋  1ݎଶ െ 14 ܮଶሺߠ ൌ 0ሻ   ሺ݈݅݊݁ ݁ܿݎݑ݋ݏሻ  g(r) is a relative dose function that accounts for the variation of dose with distance along the transverse axis of the source due to absorption and scatter by the medium (i.e. tissue or water). It may also be influenced by filtration of photons by the source materials. By definition, g(r=1cm) =1.The anisotropy of the dose distribution around a source (excluding geometric effects) is accounted for by an anisotropy function, F(r, θ). However, for many applications, such as when source orientation is not known or not predictable, anisotropy is approximated by an anisotropy factor, φan(r).   15The various TG43 parameters and the exponential decay equation for the radioisotope in question are combined to calculate the dose rate in water due to a single seed:  ܦሶ ሺݎ, ߠ, ݐሻ ൌ ܵ௞݁ିఒ௧߉ܩሺݎ, ߠሻܩሺݎ଴, ߠ଴ሻ ሶ݃ሺݎሻܨሺݎ, ߠሻ  Where λ is the decay constant of the radioisotope (λ=ln2/T1/2, where T1/2 is the half-life). In multiple seed treatments, the dose contributions from each seed are determined and summed for each dose calculation point. The total dose delivered is calculated based on the treatment time, T:  ܦሺݎ, ߠ, ܶሻ ൌ න ܵ௞்଴ ݁ିఒ௧߉ܩሺݎ, ߠሻܩሺݎ଴, ߠ଴ሻ ݃ሺݎሻܨሺݎ, ߠሻ ൌܵ௞ߣሺ1 െ ݁ିఒ௧ሻ߉ܩሺݎ, ߠሻܩሺݎ଴, ߠ଴ሻ ݃ሺݎሻܨሺݎ, ߠሻ  In the case of permanent implants (T=∞), this reduces to:  ܦሺݎ, ߠ, ܶሻ ൌܵ௞ߣ߉ܩሺݎ, ߠሻܩሺݎ଴, ߠ଴ሻ ݃ሺݎሻܨሺݎ, ߠሻ  1.2.3.4 Activity  Activity, A, is defined as the rate of nuclear disintegration or transformation within a radioactive source. Its unit is the Becquerel (1 Bq = 1 disintegrations/s). We will freely use the more traditional unit, the curie (1 Ci = 3.7× 1010 disintegrations/s=3.7 × 1010Bq) which is still commonly used in clinical practice. Each disintegration represents the spontaneous transformation of an atom from one nuclear state to another. For most brachytherapy radionuclides, such nuclear state transformations are accompanied by emitted photons in the form of unconverted γ–rays and characteristic x-rays, as discussed previously, as well as annihilation photons (for β+ emitters)  and Bremsstrahlung photons created by interaction of charged particle emissions with the source material. Activity is measured by counting the number of photons, β particles, etc., emitted by an unencapsulated point source of the   16radionuclide by means of scintillation or proportional counters, from which its activity is inferred 17. For sealed brachytherapy sources activity refers to that contained inside the encapsulation. Activity was used to specify source strength in older brachytherapy dose calculation formalisms, but has been replaced by the more easily measured Air Kerma Strength in the TG43 formalism.  Nevertheless, because of its long history, activity is still frequently referenced in clinical practice.  For low energy sources, such as I125 and Pd103, which suffer from significant attenuation by the source encapsulation, source strength may be described using the apparent activity, Aapp. Aapp is defined as the activity of a hypothetical unfiltered point source of the same radionuclide that would give the same air-kerma rate in air at a reference distance (typically 1 m) along the perpendicular bisector of the source.    1.3 Treatment Planning for Permanent Implant Prostate Brachytherapy  1.3.1 Target Volume Definition  Identification of the target volume in prostate brachytherapy initiates the treatment-planning process. Most commonly, an ultrasound volumetric study of the prostate gland is obtained for target delineation a few weeks prior to the operative procedure. The volume study is acquired with the patient in the dorsal lithotomy position using a transrectal ultrasound probe mounted securely to a mechanical stepper and stabilizer unit. Images are acquired at 5 mm intervals, starting from the proximal seminal vesicles just superior to the bladder neck/base of the prostate gland and extending to the apex.  The prostate, and sometimes a small amount of neighboring tissue, is identified as the clinical target volume on the ultrasound images. A margin to account for targeting uncertainty is then added to create the planning target volume, or PTV. Typical margin settings are 4 to 6 mm laterally and anteriorly, 0 mm posteriorly, and 6 to 7 mm superiorly and inferiorly. Depending on the planning approach used (see following section), the resulting PTV is used to either create a treatment plan, or to estimate the number of seeds that will be required for the implant.     171.3.2 Planning Approaches  The table below lists the type of planning applicable to prostate brachytherapy as defined by the American Brachytherapy Society (ABS).18   Planning approach Definition Preplanning Creation of a plan outside the operating room (OR) hours, days or weeks before the implant procedure. Intraoperative Plan created in the OR. The patient remains stationary between the time of the volume study and implant procedure. Interactive The treatment plan is revised periodically during the implant procedure using image-based feed-back of needle position to recalculate dose. Dynamic dose calculation Dose distribution continuously updated using deposited seed position feedback. Table 1: American Brachytherapy Society (ABS) Nomenclature for Different Types of Planning Used in Permanent Prostate Brachytherapy  All of the above methods use transrectal ultrasound imaging and mechanical template guidance during the implant procedure (figure below).  Seed placement is also frequently monitored with fluoroscopy, as seeds can be difficult to visualize on the ultrasound images. A grid corresponding to the template holes is superimposed on the ultrasound images to aid in positioning needles during treatment planning and implantation. Once finalized, the treatment plan will designate the required needle positions according to the holes in the guidance template, and the location of each seed in each needle relative to the needle tip.   18 Figure 3: Stepper system and template  Seeds are delivered to the predetermined locations in the prostate under transrectal ultrasound and template guidance by one of the two methods. The first method involves the use of preloaded needles; the seeds are loaded into the needles in locations specified by the plan prior to insertion of the needles into the tissue. The second method involves the Mick applicator system, which allows seeds to be deposited one at a time into needles after they have been inserted into the tissue. The Mick applicator procedure takes longer but offers more control in terms of adjusting seed spacing, which can be monitored during the procedure with fluoroscopy. A new development by Nucletron is the Seed-Selectron which is used for prostate implant. The system comprises a small pre-ordered source cassette for the seeds that are required for one patient, with another cassette for the spacers. For each needle, the Seed-Selectron must be connected and the number of sources calculated in an on-line treatment planning program can be prepared: an individualized train of sources and spacers is then composed. The source train is pushed mechanically through the needle, thereby passing a small detector, which verifies the strength of each individual source. When the sources are positioned in the prostate and the push wire is retracted, the process can be repeated for the next needle.    19Two of the planning approaches listed in Table 1 are in routine use today. The most common approach involves pretreatment planning with TRUS, although intraoperative planning is gaining in popularity. Combinations of preplanning and intraoperative planning are also in use. The preplanning approach requires significantly less time in the operating room, as the treatment plan and all materials are prepared in advance on the basis of the pre-operative ultrasound study, but offers less precision because the implant occurs a few weeks after the planning images are acquired. Patient positioning, as well as the shape, size, and position of the prostate on the ultrasound images acquired at the time of implant may differ slightly from their appearance on preplanning images. Achieving acceptable dosimetry can sometimes be a challenge under these conditions.18 In contrast, intraoperative planning is done on the basis of the intraoperative ultrasound images. In addition, feedback on dosimetry (based on information about actual, versus planned, needle targeting) can be obtained during the implantation procedure, and the plan adjusted accordingly. However, there are also drawbacks. Planning is done under significant time constraints in the operating room, limiting the ability to critically review and double-check the plan. There can be a significant waste of unneeded seeds at the time of the procedure since an approximate number of seeds must be ordered in advance. If preloaded needles are used, seeds must be loaded into the needles in the operating room, requiring additional operating room time, which is costly. Because seed delivery into the tissue is not very precise, regardless of the planning technique, the advantages of intraoperative planning versus preplanning are still under debate.18  The work described in this thesis applies primarily to the preplanning approach, although it could also be useful in some intraoperative settings.   1.3.3 Needle and Seed Placement Approaches  Specialized treatment planning software, of which there are several commercially available, is used to formulate the treatment plan using the TG-43 dose calculation formalism.  Source placement patterns are obtained either through manual (forward) planning or computerized (inverse) planning. For permanent implants, plans are created using seeds with a single activity.  This is done because post-implant quality assurance requires   20calculation of the post-implant dose distribution based on seed locations derived from CT images.  Although some progress has been made in this area19, spatial correspondence between planned seed positions and their position on the CT scan is not easily established, so seeds cannot be uniquely identified relative to the treatment plan.  All seeds must therefore be assigned the same activity.   Even with uniform seed activity, there are many degrees of freedom in achieving an acceptable treatment plan. Therefore, seed and needle placement patterns within the target volume can vary significantly, but can be loosely described as follows.  1.3.3.1 Uniform Loading  In the uniform seed loading approach, large numbers of low strength seeds are evenly distributed, usually with ~1 cm cubic grid spacing, throughout the target volume. This approach was established in the earliest days of brachytherapy. However, it results in peripheral scalloping of the 100% isodose, and very high central dose because of cumulative effects. In the prostate, the high central dose region covers the intraprostatic urethra with doses that can exceed 300% of the prescribed dose as can be seen in figures below.   21 Figure 4: DVH report of uniform loading      22 Figure 5: Uniform loading treatment plan   23Unacceptably high urinary morbidity led the Seattle group, who were amongst the pioneers of modern prostate brachytherapy techniques, to abandon uniform loading in favor of a modified version within 2 years after the start of their program.20, 21   1.3.3.2 Modified Uniform Loading  Modifications to the uniform seed loading included placement of additional needles on the periphery to reduce scalloping, and to remove some of the central seeds by removing some of the central needles, or by removing some of the central seeds from these needles.   1.3.3.3 Peripheral Loading  Peripheral loading is a further refinement of modified uniform loading, with less emphasis on uniform seed spacing. An ideal peripherally loaded prostate has all the sources 5 to 10 mm inside the prostate capsule to spare the rectum and bladder while providing an adequate dose to the prostate plus margin. Judicious placement of sources on the periphery of the target volume can cause interior doses above the prescription threshold and below the level that will damage the urethra. Nevertheless, peripheral loading is, like uniform loading, as an idealized formalism that, in practice, is routinely modified rather than executed in its pure form. Although no definitive recommendations have been proposed for pre-implant prostate dosimetry, AAPM TG-56 suggested that treatment plans be “designed to place seeds peripherally to improve dose homogeneity and to avoid unnecessary radiation damage to the urethra”.21 Treatment plans generated at our institution have features of both modified uniform and peripheral loading. An example is shown in figure 8. Needle position is designated according to the template grid position, and a diagram of seed positions within each needle is generated by the treatment planning system (VariSeed – Varian Medical Systems). (Figure 6 and 7)   24 Figure 6: Needle loading report     25 Figure 7: Seed position report      Figure 8: Modified loading treatment plan 26   271.4 Brachytherapy Source Train Assembly and Needle Preparation  As illustrated in figure 9, radioactive seeds for permanent implantation are loaded into needles as source-spacer trains of variable length, where the radioactive seeds are separated by one or more non-active spacers. A typical treatment plan can have between 10 and 40 needles loaded with 50-150 seeds.  Spacers are made from bio-absorbable suture material, and come in various lengths.  The most commonly used spacer length provides 1 cm center-to-center seed spacing if seeds are separated by a single spacer. (The exact length of the spacer will depend on the length of the seed). Seed-spacer trains with 1 cm center-to-center seed spacing are referred to as “standard” loads, while other spacings are called “special” loading.  Figure 9: example of seed configurations    The radioactive seeds and spacers are inserted into the needles as either loose (unconnected) trains, or as connected strings or strands, achieved by encasing the source-Regular 1 cm spacing ‘’standard load’’ Examples of modified spacing ‘’special load’’ Adjacent seeds  (rarely used)   28spacer trains in a tube of bio-absorbable material (figure below).  Stranded seeds are preferred by many end-users as they offer clinical advantages such as reduced seed migration and loss, and, consequently, reduced embolization to the lung.22    Figure 10: Example of a stranded seed train prior to insertion into the needle  Brachytherapy seeds used for permanent implantation must be sterilized prior to use.   If loose seed trains are required, the seeds can be sterilized via autoclave (steam sterilization) before assembly and loading, as spacers and needles can be purchased already sterilized.  In this event, source train assembly and needle loading must be done under sterile conditions.  Stranded seed trains are typically sterilized after the stranding step, or after the needles have been loaded, as the stranding process is not performed under sterile conditions.  Due to the nature of the stranding materials, sterilization methods using heat and/or moisture cannot be used, and the most suitable alternatives, such as gamma irradiation, are not readily available in the clinic. Two options are available for the end user: purchase sterile strands consisting of 10 seeds separated by spacers to give 1 cm seed spacing, or purchase custom configurations that have been loaded into needles and sterilized.  The second option is also available for loose seed loading.  While the first option is less costly, it requires the 10-seed strands to be cut into the lengths required by the treatment plan and loaded into needles aseptically by the end user. The stranded seed-spacer configuration is also limited to the uniform 1 cm option.   29There are a number of complexities involved in loading needles by an end user. If loose seeds and spacers are used, considerable numbers of seeds and spacers have to be handled individually. The seed/spacers are small e.g. typically 4-6 mm long, and this leads to time-consuming manual handling under aseptic conditions, with an associated radiation dose hazard. Due to the complexities involved and the time-consuming nature of the needle- loading process, the preloaded needle option is increasingly desirable because it provides needles loaded to prescription with either stranded or loose seed trains and delivered sterile for immediate clinical use. Currently, a service for providing preloaded needles involves an end user placing an order from a seed manufacturer according to the prescribed treatment plan (i.e., number and type of seeds). The seed manufacturer then fills the order from its inventory. Most manufacturers send the required number of seeds and spacers to a third-party needle loading facility, although some have in-house capabilities. After loading the seeds, these preloaded needles are sterilized. Finally, the labeled and packaged preloaded needles are shipped to the end user.     1.5 Calibration of Brachytherapy Sources  1.5.1 Calibration Methods  The types of brachytherapy sources considered in this work emit photons only, as they have encapsulation that stops any charged particle emissions that may be present.  In the clinic, sources of this type are calibrated using a well, or re-entrant, chamber, which is an ionization chamber with sufficient collection volume to obtain a reasonable signal from LDR brachytherapy sources, which have very low air kerma rates. This device has cylindrical geometry, where both the inner and outer electrodes are hollow cylinders, and the radiation source is placed in an air space inside the inner cylinder for the ionization measurement (figure below).   30  The response of a well chamber to a radioactive source is dependent on the geometric position of the source in the chamber. Well chambers are therefore supplied with specially configured source holders that place the brachytherapy seed in a defined and reproducible position within the chamber. Under these specific geometric conditions, a calibration factor that relates ionization readings to radioactive source strength may be determined. Such factors are obtained by sending the chamber and its read-out electrometer to an Accredited Dosimetry Calibration Laboratory, which determines calibration factors for different brachytherapy source models on the basis of comparison to known standard sources provided by NIST (the National Institute of Standards, USA).  NIST performs absolute calibration of its primary standards using a Wide Angle Free Air Ionization Chamber (WAFAC).23   1.5.2 Quality Assurance Standards for Source Strength Verification  A common clinical brachytherapy dosimetry practice of the past was to accept the manufacturer-specified source strength without verification as a basis of treatment planning. collecting electrode source holder to electrometer outer electrode insulating material Figure 11: Schematic diagram of a well type chamber   31But it is an unwise approach and AAPM Task Group 4024 stated that ‘Each institution planning to provide brachytherapy should have the ability to independently verify the source strength provided by the manufacturer.’ Manufacturers deliver sources within a specified range of the requested nominal activity; however, source strengths can sometimes fall outside this range, or the mean strength deviates significantly from the requested nominal value. The medical physicist is responsible for the dose given to the patient, so the medical physicist should independently validate the output of seeds using appropriate equipment with secondary traceability to a National Standards Laboratory (i.e. a well chamber with calibration factors obtained from an ADCL for all source types being used). As validation of all seeds can be impractical for LDR implants utilizing a large number of seeds, such as prostate brachytherapy implants, the AAPM Task Group 40 recommended that a minimum of 10% of the seeds be verified in such cases. For smaller orders, ≥10 sources should be assayed, and for orders of <10 sources, all sources should be assayed. If more than one strength grouping is present in an order, ≥10% or ten sources, whichever is larger, of each strength grouping should be assayed and all the sources in strength groups with < ten sources. In subsequent publications by AAPM Task Groups 56, 64 and 98 21, 25, 26 , it is recommended that if the institution’s verification of source strength disagrees with the manufacturer’s data by more than 3%, the source of the disagreement should be investigated. They further recommend that an unresolved disparity exceeding 5% should be reported to the manufacturer.   1.5.3 Statistical Considerations  Consideration of sampling statistics can aid in interpretation of seed assay results in the context of the AAPM recommendations. For assay of the mean source strength, a detailed analysis of the statistics of source assay has been presented by Yue et al. 27 In this analysis,  it was found that the number of sources from a typical permanent seed prostate implant order that must be assayed to assure that the measured sample mean differs from the population mean by less than the recommended 3% tolerance threshold is dependent on the measured deviation of the source strengths. For example, they calculated that for a sample strength   32deviation of 5%, 14 sources must be assayed to determine that the sample mean is within 3% of the population mean with 95% certainty. This number is larger than 10% of the sources for shipments of less than 140 seeds, but is less than 10% for larger shipments.  Definitive use of such an analysis would require knowledge of the true strength deviation as well as the shape of source strength distributions as a function of radionuclide and manufacturer, which may not always be available. However, estimated deviations, like the 5% value employed above, can be of assistance in situations where the recommended 10% of seeds cannot be assayed, or when results do not meet the AAPM recommended tolerance. For assay of individual source strength, seed strength distribution must also be considered. A known example exists for manufacturers of I125 sources that employ a source binning strategy where sources are assigned to “weekly” activity bins. These bins are nominally 8% wide (e.g., ±4% tolerance) because I125 decays about 8% per week.28 Measurement uncertainty in assigning sources to the appropriate bin is expected to make the actual bin width somewhat greater. Clinical measurements of seed strength distributions reported in the literature29-31, are consistent with a 5% value for two standard deviations at the 95% confidence level. If the sample mean is allowed to differ from the manufacturer’s value by 3% (which can occur due to combined calibration-coefficient uncertainties between the manufacturer’s and user’s instruments), individual sources will exceed the 5% threshold more than 5% of the time. Combining in quadrature the ±5% spread in seed strengths with the 3% mean tolerance results in a 6% individual seed strength tolerance, which is slightly greater than the tolerance recommended by the AAPM for individual seeds.28  1.5.4 Calibration of Sources in Sterile Needles, Cartridges, or Strands  The TG-40 and TG-56 reports recommended the purchase and calibration of a single loose source from each strength grouping when the remaining sources intended for implantation were in sterile strands or assemblies. This criterion was considerably weaker than the 10% recommendation for loose, nonsterile sources, because there was no established method for assaying sources in sterile assemblies at the time, and sources not implanted into the patients represented additional cost that could not always be accommodated.  Clearly, this low rate of sampling would result in frequent instances of readings outside the recommended   333% mean tolerance, and even the 5% individual seed tolerance. It is also possible that the assay seed does not come from the same lot as the stranded seeds, as source manufacturers who also sell stranded source products do not always guarantee that loose sources supplied with an order of stranded sources will come from the same batch because of logistics within the factory.28 This situation has been alleviated by techniques and commercially available hardware that has been developed for calibration measurements of sterile strands, cartridges, or pre-loaded needles.  These exist in the form of specialized well chamber inserts that can be used in conjunction with the ADCL-measured single-source calibration coefficient for the well chamber. 32, 33 Similar source assay techniques had been published,34, 35 and acknowledged in the TG-64 report of 1999, but were of limited use because of noncommercial source holders and pronounced source positioning effects that have since been addressed in the commercially available versions. Even so, these current methods of assaying sterile assemblies are limited, as all seeds in the assembly are assayed simultaneously, which yields only a bulk reading for the entire strand. In addition, bulky equipment in the form of the electrometer, well chamber, and sterile inserts must be deployed in a sterile setting relatively close to the time of implant. Another option that utilizes an imaging plate to assay sterile source assemblies has been reported by Furutaniet al.36  While this group reported promising results with this approach, no further development appears to have taken place.  1.5.5 Third-party Brachytherapy Source Calibrations  Several third-party source handling and calibration services, including needle loading facilities, offer independent assays of brachytherapy source strength. Because of semi-automated processing in a well-shielded environment, third-party source handling and calibration services (henceforth referred to as third-party services) typically assay all the sources in a customer’s order prior to loading into needles as either loose or stranded assembles. These assays are more comprehensive than those required of the end user by AAPM task groups; even if it was always possible for the end user to perform the recommended calibration checks on the sterile source assemblies.  Nevertheless, the AAPM   34Task Group 98 report28 suggests that exclusive reliance on the third party calibrations raises significant medical physics, patient safety, and legal issues. In summary, they question whether source strength verified by outside entities can be a reliable substitute for the final check by the end user.  For example, reliance on the third party calibration report does not guard against errors subsequent to the assay, such as the vendor misaddressing the order or mixing sources from inappropriate strength groupings after the calibration step. Consequently, the AAPM Task Group 98 revised and supplemented the recommendations of previous reports to produce the guidelines summarized in the following table.   Source Form Number to be assayed*All loose sources, nonsterile ≥10% of total or 10 seeds, whichever is larger. Nonsterile cartridges ≥10% of total via whole cartridge assay or via single sources. Mixture of nonsterile loose sources  and sterile assemblies Loose sources amounting to ≥10% of the total order or ten seeds, whichever is larger. Sterile source assemblies ≥10% of assemblies via sterile well chamber inserts or quantitative image analysis. Alternatively, order and assay nonsterile loose seeds equal to 5% of the total or five seeds, whichever is fewer. Strands ≥10% of total or two strands, whichever is larger, using single-seed calibration coefficient.37 Alternatively, order and assay nonstranded loose seeds equal to 5% of the total or five seeds, whichever is fewer. * Each source-strength grouping in an order should be sampled. If the number of sources in a strength group is <10, the entire group should be assayed. Table 2: Quantities of brachytherapy sources to be assayed by the end-user physicist  Further, it is recommended that the mean source strength value to be used in treatment planning be either that stated on the manufacturer’s certificate or the value determined by the institutional medical physicist when the two values agree to within 5%. Individual source outliers may be discarded or retained at the discretion of the medical physicist and radiation oncologist. It also recognized that the range of source strengths in a given order is manufacturer and radionuclide dependent, and therefore allows some discretion in application of the 5% action threshold for a single source.   35If sources are ordered for the purpose of end user assay, preloaded needle vendors will forward these sources from the same lot used to load the sterile product.  They are sent in a shielded vial along with the preloaded needle package.  1.6 Purpose of this Thesis  Seeds containing the radionuclide I125 are the ones most commonly used for permanent implant prostate brachytherapy. Seeds are custom ordered for each implant, and there is a small probability that a seed shipment will arrive at the clinic with the incorrect source strength, in spite of quality assurance steps taken by the end user during the ordering process, and by the supplier in preparation and delivery of the sources. Medical physics practice recommendations therefore advocate that a final check of the source strength be performed by the end user. Equipment and procedures to perform these checks are well established for brachytherapy sources that have not yet been loaded into needles, but not for sources that arrive in sterile, preloaded needles. For the latter option, current technology limits the verification to a small sample of seeds that is not part of the preloaded needle set, or to bulk assay of multiple seeds in the needles. The aim of this work is to develop a method to verify, under sterile conditions, the source strength of individual brachytherapy seeds already loaded into needles, thereby enabling an end-user quality assurance check on 100% of the seeds to be used in an implant.              36Chapter  2  Materials and Methods  2.1 I125 Sources   Two commonly used I125 seed models, Oncura 9011 and 6711 (GE Healthcare, Medi-Physics Inc., Arlington Heights, IL) were used for this work. The model 6711 seed contains a silver wire with the active material, silver iodide (AgI), adsorbed on its surface (figure below). In this design, the silver wire is present to assist in visualizing the seed position and orientation on radiographs. The silver substrate adds characteristic x-rays from silver to the I125 spectrum. These characteristic x-rays lower the average energy of the emitted spectrum.38   Figure 12: Schematic diagram of I125 seeds, Model 6711   The model 9011 is similar to model 6711 but it is thinner, with a 0.51 mm outer diameter.  It can therefore be loaded into thinner (20 gauge rather than 18 gauge) needles.  2.2 Radiation Detector  We investigated a method using a flat panel diagnostic x-ray image detector to act as a dosimeter to characterize the individual source strength of seeds loaded into brachytherapy needles. In using an imaging detector, the dosimetry method explored has some similarities to the one described by Furutani et al.36  Provided it has suitable characteristics for the   37application, a diagnostic energy (kilovoltage) flat panel image detector can act as a dosimeter in a similar way to an imaging plate (as used by Furutani et al), film39, or electronic portal imaging devices (megavoltage flat panel detectors) 40. Flat panel detectors have several advantages over alternative technologies, such as film or older digital imaging technologies.  These advantages include: real-time image read-out, image processing/enhancement options, large active area, high frame rate, low noise, high spatial resolution, and linear dose response.  2.2.1  Flat Panel Detector Technology  Flat panel detectors convert an incident x-ray signal into a digitized image.  They are constructed of multiple components that include, at a minimum, an array of electronic photon sensors (photon-electron converters) and an array of read-out electronics. Most flat panel detectors designed for x-ray imaging applications also include a layer of scintillating material placed over the electronics, which converts the x-ray photons to visible light photons.  This is called indirect conversion, and enhances the detection efficiency, since most photon sensors are more sensitive to visible light than to x-rays.  Commonly used scintillating materials include cesium iodide (CsI) or gadolinium oxysulphide (GOS). Early flat panel technology used amorphous silicon for the electronic components.41, 42. This technology is well-suited to many imaging applications, and is still the most commonly used today. More recent developments include the use of complementary metal-oxide-semiconductors (CMOS) as the active pixel sensors. CMOS offers advantages in the form of faster image acquisition and lower read-out noise, which is advantageous for some types of imaging applications.43. A schematic of such a detector is shown in the figure below:     38 Figure 13:  Schematic diagram of CMOS detector      X-rays Scintillator Microlens optics CMOS sensors Controller electronics Fiberoptic plate X-rays Scintillator  (X-ray to optical photons) Optical Coupling CMOS   392.2.2 Choice of Flat Panel Detector   None of the commercially available detectors are designed specifically for our purpose, as the commercial focus is on radiologic image detection or industrial applications rather than dosimetry. Our goal was therefore to identify the detector that was best suited to our application. In assessing the products available, we focused on the following technical criteria: 1. Spectral sensitivity: for detection of I125 emissions, we needed a detector designed for optimal detection of x-rays in the 10-50 keV range. 2. Detection efficiency (speed) and selectable integration time: LDR brachytherapy sources emit very low intensity radiation, so integration times had to be highly selectable and high detection efficiency was required to keep data collection times reasonable. 3. Spatial resolution: the autoradiographical images produced by individual brachytherapy sources are small, so high spatial resolution was required. 4. Detector size:  the detector had to be large enough to collect images from a 6-7 cm long seed-spacer train (the longest typically encountered for prostate implants) and, preferably, to accommodate several needles spaced at 1-2 cm intervals. 5. Interface: the interface with display and control software had to allow flexible user control of all the imager’s features. Most commercially available detectors are designed for photons of energy greater than 40 KeV, the exception being mammography x-ray detectors. Amongst the mammography detectors available, we chose the Dexela Model 1512, which was the most suitable in terms of the other criteria. Dexela produces a number of flat panel CMOS X-ray detectors for a range of applications including mammography, breast tomosynthesis, fluoroscopy, dental Cone Beam CT, small animal imaging and non-destructive testing. The model 1512 detector has a sensor unit consisting of 1536 columns × 1944 rows with a 75 μm square pixel pitch. The sensor can operate with variable binning (i.e. combining of pixels) that allows trade-off between spatial and temporal resolution.  Table 3 and Figure 14 provide additional technical details for this.    40 Resolution 1536 × 1944 pixels Pixel sampling resolution 14 bits Pixel pitch 74.8 µm Active image area 145 × 115 mm Exposure integration time 38 ms to 10 s DQE with 600 µm CsI scintillator: RQA5 : 74 kVp, 21mm aluminium>0.7 at 0.5 LP/mm (1×1 binning, high full well mode) Read out format 6 parallel channels Interface Camera Link Base configuration Power dissipation 11W (active), 0.5 W (standby) Operating temperature range +10C to +40C Max frame rate (Camera Link) 1×1 (no binning) 26 fps 2×2 70 fps 4×4 86 fps Max frame rate (GigE) 1×1 (no binning) 11 fps 2×2 46 fps 4×4 86 fps Optimal energy detection energy 30-250 KeV  Table 3: Technical data 1512      41 Figure 14: Enclosure Dimensions (1512)  As in most flat panel designs, the CMOS sensors are optically coupled via a fibre optic faceplate (FOP) to a scintillator plate which is specifically designed to convert x-ray photons in the energy range 30-250 keV to light photons. In addition to the optical coupling, the FOP shields the sensor from a significant amount of x-rays, which can otherwise generate image noise and cause gradual damage to the sensor. The scintillator, composed of Cesium Iodide, is pressed in contact with the FOP but not glued. The unit is housed in an aluminium enclosure, with a carbon fibre composite panel providing an x-ray ‘window’ in the image field of view. The unit housing also acts as a heat sink, as the detector has no cooling fan.  A circuit board (ADC) connected to each sensor digitizes the analogue signals, and provides the timing signals needed to drive the sensors. A communication board (DAQ) converts the data stream from the ADC(s) to the Camera Link. Camera Link is a hardware specification that standardizes the connection between cameras and frame grabbers. It defines Max.241223.5150 3114515 152Sensor Area 42 18.25 53 170 M4×0.7 Fixation Threads   42a complete interface which includes provisions for data transfer, detector timing, serial communications, and real time signaling to the detector.  2.2.3 Imaging System Set-up   For the user interface with the image detector, we installed an EPIX E4 frame grabber board into a spare PCI Express slot in the host PC and then installed the device drivers. The DC power cable was connected between the power supply unit (PSU) and the detector and also the PSU mains cord to the outlet. The Dexela software was installed, and the camera link cable was connected between the detector and the EPIX board in the PC.  The detector is a programmable data acquisition system with internal memory. It communicates with the host computer using a combined data and control link, which is Camera Link as described previously. Each detector is factory calibrated for optimum performance, and the calibration settings are held in a non-volatile memory in the detector. SCap is a utility for driving the Dexela series of x-ray detectors. It can be used for demonstrating, testing or evaluating the detector, or as a stand-alone x-ray work station. SCap handles images in two formats: a 16 bit fixed point format, which contains the 14-bit pixel values (0 to 16383) and a 32 bit floating point format. The 16 bit format is used for most images: captured images, dark current images and corrected images.  There are several options in the control panel of the SCap that can be selected: Descramble: Converts the raw interleaved image from the detector into a legible format. This option is on by default. Auto-contrast: Automatically adjusts the brightness and contrast of each image, so that the grey levels are expanded to the range of the display. Dark correction: Automatically performs dark correction on each captured image. Correction is only performed if there is a dark image available. The dark image is an average of several dark frames. Dark correction removes static offsets between pixels. Gain correction: Automatically performs gain correction (flood field correction) on each captured image. Correction will only be performed if a flood image is available.   43Linearization: Applies a software linearization to each image captured. Detector pixels have a slightly non-linear response to light, which corresponds to about 2% deviation from the linear response at mid-scale, so the manufacturer recommends that this option be left on. Defect correction: Applies a masking function to defective pixels in the image sensor. Some defective lines and pixels may be present on the sensor. These are defined as having a response which deviates by a specified amount from the ideal. A defect map is created during factory testing of the detector and is installed in the SCAP program folder. On the final processed image, pixels tagged on this defect map are replaced by interpolated values in order to improve the image presentation.  Suppress zingers: Removes isolated white pixels which may appear from direct conversion of x-ray photons in the CMOS sensor. SCAP can provide automatic correction of x-ray images immediately after they are captured. We applied all these corrections to the captured images except the gain correction because we couldn’t take a flood field with the x-ray sources that we had.  Flood fields are meant to be acquired using a typical imaging geometry (i.e. a single x-ray source at large distance) and are therefore not applicable to our application (multiple small sources close to the detector surface).  The dark image acquisition settings can be entered in the configuration menu: exposure time, number of averaged images and, optionally, the path for the dark image file. A new file is created each time a dark current image is acquired; old files are not automatically overwritten. We can select and load the desired dark image from the files. We can also enter the exposure time per frame in milliseconds (minimum 37ms).  Nine on-sensor pixel binning modes, as well as off sensor binning at 2 × 2 are available if required.  The detector is designed for maximum sensitivity to x-rays in the specified energy range (30-250 KeV), enabling optimal image quality for a low dose exposure. It is recommended to use a shorter detector exposure time, and to capture more than one frame. This will generate an average of several frames, rather than a single one. The averaging process reduces the image noise. The number of averaged frames can be set. Two settings of pixel gain can be programmed in the SCAP: High Full Well mode is recommended where high dynamic range is important - this applies to most static x-ray   44imaging applications. Low Full Well mode increases the pixel gain by approximately 300%, and can be used where high sensitivity (low dose per frame) is required. Five modes of exposure are available: Normal mode: A single frame is exposed after a trigger (Snap button or external trigger). Pre-shot mode: Sets the detector into an external edge mode sequence of two exposures, usually one short and one long one. When the exposure is finished the detector reverts to Normal mode for the next exposure. Pre-programmed Exposure mode: A sequence of up to 4 frames is exposed, each frame with an individually programmable exposure time. The exposure time for each frame is set in a dialog box which appears when this option is clicked. This option can be used for making composite exposures where the subject has a high density ratio. Frame rate mode: A series of exposures is captured with a delay (minimum 38ms) between frames. The number of frames, delay time and exposure time are set in a dialog box which appears when this option is clicked. Software trigger mode: the detector is triggered by the software and images are acquired under SCap control. The Darks, Floods, Average/Sequence collection modes use this trigger. The Software trigger mode was used to capture all images for our study. We performed the following steps to calibrate the detector: The default configuration file was loaded after launching SCap. In the configuration menu, Low Full Well and 1 × 1 binning were selected. The exposure mode was set to normal and the software trigger mode was used. A single exposure was used to capture a dark image.   2.3 Imaging Geometry  The imaging technique used in this work, and the related work of Furutani et al36, falls into the category of autoradiography. An autoradiograph is an image produced by the emissions from a radioactive source or sources. Autoradiographs are almost always acquired with the image detector as close as possible to the sources (“contact” autoradiography), and are used primarily as a means of localizing the radioactivity. Examples of routine use of autoradiographs in brachytherapy include source positioning checks for HDR afterloaders, and needle loading pattern checks of pre-loaded brachytherapy needles. Third-party needle   45loading facilities routinely send contact radiographs of the needles with each patient –specific shipment. An example is shown in figure below.  Figure 15: example of needle autoradiograph  Placing the radioactive sources close to the image detector provides the best possible spatial resolution. However, for a dosimetry application, where image intensity is used to measure dose, this short distance causes difficulties due to inverse square fall-off effects, as described below.  2.3.1 Geometrical Factor Uncertainty  Dose fall-off due to the inverse square law, which is described by the geometry factor in the TG-43 dose calculation formalism, is extremely steep at clinically relevant distances for brachytherapy (i.e. on the order of 1 cm), and is even more extreme at the sub-millimeter distances between source and detector in a contact autoradiograph. Thus, any uncertainty in the distance between source and detector can cause extremely large uncertainties in the image intensity, and, therefore, the dose measurement derived from it. Figure below shows the effect of a ±0.1mm uncertainty on the geometry factor at different distances from the source. Values were calculated from r=0.04cm to r=1.00cm and with 0.02cm increment, using both point and line source approximations and the discrepancy was defined as: ଵଶ ൈ  ሾܩሺݎሻ െ ܩሺݎ ൅ 0.1ሻܩሺݎሻ൅ ܩሺݎሻ – ܩሺݎ െ 0.1ሻܩሺݎሻሿ   46r was calculated along the perpendicular bisector of the source (θ=90o in the TG-43 formalism). The seed was assumed to be centered on the origin and the seed active length was 3mm, which is the accepted TG-43 value for the I125 sources used in this study.15   Figure 16: Discrepancy of geometrical factor for 0.1mm distance uncertainty  At points close to the source this discrepancy is extremely large, but decreases rapidly with distance.       0%10%20%30%40%50%60%0.00 0.20 0.40 0.60 0.80 1.00distance (cm)discrepancy for 0.1 mm distance uncertaintyline sourcepoint source  47Factors that can contribute to geometric uncertainty in an autoradiograph of preloaded brachytherapy needles include: 1. variations in needle outer diameter and wall thickness 2. variations in seed position within the diameter of the needle lumen 3. effects due to bowing of the needle (i.e. poor contact) 4. thickness variations in any material placed between the needle and the image detector in order to maintain a sterile field 5. variations in the distance of the image sensors from the surface of the detector Assuming that (3) can be avoided by placing appropriate pressure on the needles, and that (4) is negligible, the main sources of uncertainty for an autoradiograph is due to the needle (1 and 2) and the image detector (5). Information about the needles used for permanent prostate brachytherapy in our institution was obtained from the supplier and is shown in table below.  Needle Outer Diameter Inner Diameter Wall Thickness Gauge Nominal (mm) Tolerance ±(mm) Nominal (mm) Tolerance ±(mm) Nominal (mm) Tolerance ±(mm) 18 1.270 0.013 0.838 0.038 0.216 0.013 20 0.9081 0.0064 0.6030 0.0190 0.1524 0.0064 Table 4: 18 and 20 gauge needles information  Using these parameters, a “worst case” seed position uncertainty relative to the detector surface (Table 5) can be estimated, based on the assumptions in the figure below.    “nominal” position” δ=0 “closest” position (thinnest wall) “farthest” position (thinnest wall and largest outer diameter) “-”  “+” reference surface Figure 17: Seed position uncertainty relative to detector surface   48Needle Seed distance from reference surface (mm) δ (mm) Gauge Model diameter (mm) nominal “closest” “farthest” “closest” “farthest”18 6711 0.8 0.635 0.603 0.680 -0.032 0.045 20 9011 0.51 0.4539 0.401 0.5135 -0.0529 0.0596 Table 5: Seed position uncertainty relative to detector surface  It should be noted that while the scenario for the “farthest” may seem unlikely, it could occur due to bunching of loose seeds and spacers within a needle, or due to inherent curvature in a stranded seed train. Since information about the detector’s geometric tolerances is not available, we will assume they are of about the same order of magnitude as the uncertainties due to the needle tolerances – in other words, the combined uncertainty would be on the order of ±0.1 mm.  Assuming a contact autoradiograph would be obtained at a distance of 0.5-0.7 mm (i.e. allowing for a sterile sheet, slightly more than the nominal distance from the reference surface in the table above), image intensity could vary by up to 40% for point source and 25% for line source.   Thus, it is clear that steps had to be taken to reduce the geometric effects.  The only way to do this was to increase the distance between the seeds and the detector, which introduced additional challenges in the form of reduced signal intensity (a potential problem because the output of the sources is already low) and spreading of the radiation over a wider area of the detector (a problem because sources are close to each other in the needles). To contain the spread of the emissions from each seed, and also to introduce the required distance between the needles and the image detector, a sterilizable needle holder with collimating channels, which we will refer to as the jig, was constructed.  The jig will be described in more detail in a later section.        492.3.2 Additional Sources of Uncertainty  Additional sources of uncertainty also had to be considered in order to assess the overall feasibility of our approach, and to determine how much the geometric uncertainty had to be reduced.  The additional sources of uncertainty were: 1. Variable attenuation effects in the needle wall due to variable wall thickness 2. Variable response in different regions of the image detector 3. Variable photon and electron scattering in the collimating channels of the jig 4. Variable seed-to-detector distances at different locations in the jig Attenuation effects due to the needles could be estimated from the needle data (µ= 7.55 mm-1): Needle gauge % transmission through max wall thickness % transmission through min wall thickness % difference in reading 18 17.7 21.6 17.8 20 30.2 33.2 9.2 Table 6: Attenuation effect due to variable wall thickness of the needle  The percentage of difference in reading will obviously limit detection at the 5% level, but, as we discussed, in the context of the seeds already have been checked by the third party, and that the attenuation effects are more likely to be a random effect, mean reading would likely still be sufficiently accurate for a verification measurement. The other sources of uncertainty were unknown in advance, but they are not random sources of error and so could be corrected for by proper calibration of the apparatus.  2.3.3 Design of the Needle Holder  The first jig designed for our initial experiments was constructed in the machine shop at British Columbia Cancer Agency and is shown in the figure below. This jig was made like a wedge to allow various needle to detector distances to be tested. This jig provided needle to detector distances ranging from 1 mm to 1 cm in 1 mm intervals. Also, as determined from   50our analyses of the inverse square effects, distances greater than about 1cm offered diminishing returns in terms of reducing the uncertainties. As sensitivity of the selected image detector to the dose rates emitted by our brachytherapy sources was unknown and could not be ascertained from the manufacturer, the image detector was first acquired on a trial basis, and this jig was used for evaluation of the detector’s suitability for our application.   Figure 18: Wedged jig design  In performing our initial evaluations, detector sensitivity was tested to see if usable signals could be obtained within reasonable integration times. In addition, response of the detector was checked at various locations on the detector to see whether it was uniform enough or not.  In performing these evaluations, signal overlap between the radiation fields of adjacent seeds was seen using the above jig design, which could increase the uncertainty of seed image intensity, especially in the case of two adjacent seeds with different activities. Consequently, after it was established that a reasonable signal could be obtained at the 1 cm distance with this imager, the imager was purchased and a new jig was constructed with smaller holes and smaller collimator diameter. With this new design, no signal overlap was seen between adjacent seeds. It consists of 7 collimators with a diameter of 5mm and a height (distance from the detector surface to the needle surface) of 1.2 cm. The collimator can be seen in the figure below.  6 mm diameter collimator Holes at 1 cm spacing   51 Figure 19: Stainless steel collimator used in the experiments   Images taken with the old jig and the new jig are shown below. The signal overlap of the two seeds in the right side of the figure is obvious.   Figure 20: Image taken with old jig (Right), new jig (Left)       522.4 Image Data Acquisition  For creating dark calibration images, we set the Quantity (number of images to be taken and averaged) to be as large as possible (in our case 5) to get optimum image quality. When the Quantity is large, it will take longer to acquire an image, but the noise is reduced. As a general rule, the noise is reduced by a factor of √2 (≈ 1.4) when Quantity is doubled. We can also set the exposure time. Exposure time should be high enough to be able to detect sufficient photons and also it should be time-efficient. It is also recommended by the manufacturer to use shorter exposure times if possible. Five seconds was chosen based on the signal obtained from the brachytherapy seeds. The dark image should have approximately the same exposure time as the x-ray image, and should be taken with the same binning and well mode settings. Due to the warm-up effect in the detector, we took dark correction images right before each of the captured images. The exposure time of the captured image was the same as the corresponding dark image. After taking each image the Scap software will give you a ‘tif’ file. An example ‘tif’ file with seeds of 0.292 mCi apparent activity is shown in Figure 21. The ‘best image’ is assumed to be acquired when the seeds sit symmetrically on top of the holes of the collimator, so that photons coming from the seeds will reach the detector surface through the hole instead of being attenuated by the collimator material. However, the seeds inside the needle cannot be seen, so their exact location is unknown. To overcome this problem, needles were shifted along the grooves of the collimator by 1 mm and an image was captured in each position. Signals received by detector at each hole position vary in each image due to the changing position of seeds relative to the collimator hole. Image analysis (as will be discussed later) was done on all images. The image which has the highest intensity for all seed images was assumed to be the ‘best image’. This image was used to report the ‘reading’, as will be defined later, for each measurement.   53 Figure 21: Image of 4 seeds with apparent activity of 0.292 mCi  2.5 Seed and Needle Configurations   9011 seeds were used in stranded form, preloaded into the 20 gauge needles by the manufacturer. 6711 seeds, which were more readily available in loose form, were either preloaded in strands or manually loaded as loose seeds into 18 gauge needles. Manually loaded 6711 seeds were separated by Bio Spacer 910TM, which are 5.5 mm long to give 1 cm source-to source spacing when loaded alternately with 6711 seeds.  2.6 Seed Parameters and Verification  Activities of the seeds used in most of the experiments ranged from 0.15 to 0.3 mCi, which was somewhat less than the nominal activities (0.32 mCi) used for prostate brachytherapy at our institution, since unused seeds and needles leftover from clinical procedures were used for the experiments.  In addition, a few high activity seeds (6.65 and   548.49 mCi) were provided at no cost by Oncura for our initial experiments (i.e. for evaluation of the detector, as the sensitivity of the detector to the seeds was unknown at that time). The activity of seeds measured in a calibrated well chamber was defined as the gold standard. Seeds could be measured either before (in the case of manually loaded seeds) or after (in the case of preloaded seeds) readings were taken with the image detector.  2.7 Characterization of the System Response  The system response was characterized through a number of tests, as follows: Orientation test: In order to check for sloping in the needle design, we measured needles inserted into the jig in two different directions.     Figure 22: Orientation test set up    55Position test: Seeds were shifted inside the needle to investigate the effect of position of the seeds inside the needle. For this purpose, two stranded 6711 seeds were measured in the position close to the needle tip, then three spacers were inserted into the tip of the needle, pushing the seeds further in. Measurement was repeated and this time 6 spacers were inserted into the tip of the needle. This experiment was repeated with 3 different sets of needles loaded with stranded seeds with activities of around 0.205 mCi, 0.243 mCi and 0.283 mCi.  Rotation test: The activity of the seeds inside the needle was measured at 0, 90, 180 and 270 degree relative to a hypothetical axis to investigate how much the measurement was affected by the needle rotation. Temporal test: The activity of one needle with 4 similar seeds was measured 10 times consecutively to calculate the measurement variance of the detector over time. The needle position was the same during all 10 measurements. Adjacent seed test: This test was done to investigate the effect of signal overlap from adjacent seeds. One of the collimators in the jig was plugged with a piece of stainless steel that was made in the machine shop, specifically to fit into the collimator. Three stranded seeds were seated on the groove such that the middle seed was on top of the plugged collimator. Several images were captured and analyzed to see if any signal was detectable under the plugged collimator.   Linearity test with respect to time:  Images of the same seeds were captured using different exposure times. Exposure time was set from 500 ms to 10 s.  Calibration graph: We plotted the actual activity of the seeds as measured with a calibrated well chamber versus detector reading. This plot was considered the calibration graph of the detector. Since sensitivity of pixels all over the detector could be different, calibration graphs were generated for three different parts of the detector.  The position and rotation tests were intended to capture uncertainties in seed position within the diameter of the needle lumen, as well as effects due to variable wall thickness within a needle.      562.8 Image Analysis  Matlab software was used to analyze the captured images. A 2D median filter was applied to all images. Median filter is a nonlinear digital filtering technique, often used to remove noise. The pixel with highest intensity was then found for each seed image. Profile lines were drawn along two perpendicular axes passing through this highest intensity pixel. Starting from the sides of the image, the difference of each pixel value from that of its previous one was calculated for both profile lines. If the pixel values increase consecutively 10 times in the profile line along the direction perpendicular to the seed train, and 5 times in the profile along the direction of seed trains, the last pixel would be defined as a border (These numbers were selected based on a trial analysis). So, 4 borders were assigned for each seed image (Figure below). Pixels between these borders were selected and the average value of these pixels was calculated. The selected pixels were filtered again by accepting the pixels with values larger than the average and rejecting those with values less than the average. The average of accepted pixels was calculated and was reported as the ‘reading’ for each seed image.  Figure 23: Image analysis  Borders Profile line along the seed train Profile line perpendicular to the seed train  57Other methods of image analysis were also explored. For example, the pixel with the maximum value on each profile line was selected, 10-50 pixel on each side of that were chosen and their average value was calculated. However, since there were a lot of oscillations in pixel values in that region, the pixel with highest intensity was not always somewhere in the middle of the image and could vary a lot from one seed image to the other. This method didn’t seem to be very accurate and it was discarded.                             58Chapter  3  Results  3.1 Results  Orientation test: Seed measurements were done in a pre-assumed original direction and its reverse direction.   Figure 24: Detector reading in two different directions of a needle  The activity of the seeds are almost identical in two directions. The largest discrepancy was about 2%.        y = 0.9902x - 4E-05R2 = 0.98880.00710.00720.00730.00740.00750.00760.00770.00780.00790.0080.00810.00820.0073 0.0074 0.0075 0.0076 0.0077 0.0078 0.0079 0.008 0.0081 0.0082 0.0083OriginalReverse  59Position test: The results of the position test are shown in the figure below. There is negligible difference in reading at different locations along the needle which means uncertainty in needle wall thickness along the length of the needle is negligible.   Figure 25: Shifting the seeds along the needle  Rotation test: The result of the rotation test is shown below. As can be seen, needle rotation did not have a large effect on the signal.  Figure 26: Rotating the needle  0.00640.00660.00680.0070.00720.00740.00760.00780.0080.00820.0084Seed_1 Seed_2 Seed_1 Seed_2 Seed_1 Seed_2Needle_1 Needle_2 Needle_3Readingno spacer3_spacers6_spacers0.00640.00660.00680.0070.00720.00740.00760.00780.0080.00820.0084Seed_1 Seed_2 Seed_1 Seed_2 Seed_1 Seed_2Needle_1 Needle_2 Needle_3Reading0 degree90 degree180 degree270 degree  60Temporal test: The result of the temporal test is shown below. The mean value of all 10 measurements is shown in table below. The last column of the table, which is defined as relative standard deviation, shows the extent of variability in relation to mean of sample. Less than 2% of variability can be seen around the mean of each seed reading.  Figure 27: Measuring a needle with 4 seeds 10 times    Mean Standard Deviation STD/Mean*100 Seed_1 0.0078225 0.000138296 1.767923 Seed_2 0.0079161 0.00005.872 0.741814 Seed_3 0.0078559 0.000053061 0.675428 Seed_4 0.0079071 0.000106381 1.345386  Table 7: Temporal test result  Adjacent seed test: Images of the seeds on both ends of the strand were visible and no image was seen in between them, due to the plugged collimator. The area underneath the middle collimator was analyzed and no evidence of signal was seen there, showing  that the photons from adjacent seeds didn’t contribute each other’s signal.  0.00720.00730.00740.00750.00760.00770.00780.00790.0080.00810.00821 2 3 4 5 6 7 8 9 10Reading  61Linearity test with respect to time: As shown in the figure below, a linear relationship between detector readings and exposure time was observed. Detector Reading increases linearly with time.   Figure 28:  Changes in detector reading resulting from different exposure times  Calibration graph: The calibration graph obtained for the detector is shown in Figure below.    Figure 29: calibration graph for three different regions of the detector y = 7E-07x + 0.0046R2 = 0.996700.0020.0040.0060.0080.010.0120 2000 4000 6000 8000 10000 12000Time(ms)Readingy = 77.352x ‐ 0.3522R² = 0.984y = 78.924x ‐ 0.371R² = 0.9935y = 78.791x ‐ 0.3607R² = 0.99690.10.150.20.250.30.350.40.006 0.0065 0.007 0.0075 0.008 0.0085 0.009 0.0095Activity (mCi)Detector ReadingRegion_1Region_2Region_3  62The calibration graph obtained for the detector was used to predict the activity of 20 seeds, and these values were then compared to the activity measured by a calibrated well chamber. The table below shows the result. The discrepancy in the measurements appears to be random. Adding more data points to the calibration graph, with data spreading over larger range, could potentially reduce the discrepancy. Alternatively, a known reference seed that has been double checked in a well chamber could be used to normalize the reading obtained during a given measurement session.    These measurements were done with 6711 seeds loaded into 18 gauge needles. Although based on the attenuation data provided in Table 6, we expect better result for the 9011 seeds which are loaded into 20 gauge needles.  Activity measured by well ion chamber (mCi) Activity estimated by calibration curve (mCi)Percentage of discrepancy (%) 0.321 0.3152 1.81 0.317 0.3179 -0.29 0.178 0.1807 -1.51 0.179 0.1833 -2.39 0.198 0.1909 3.6 0.211 0.2033 3.64 0.238 0.2265 4.82 0.247 0.2354 4.68 0.287 0.2802 2.38 0.278 0.2727 1.92 0.284 0.2828 0.42 0.282 0.2796 0.86 0.272 0.2817 -3.57 0.267 0.273 -2.23 0.153 0.1479 3.36 0.153 0.1553 -1.53 0.252 0.2522 -0.08 0.258 0.2579 0.027 0.193 0.1963 -1.7 Table 8: Comparison of the seeds activity calculated by two different methods   63 Figure 30: Activity measured by well ion chamber vs. activity estimated by calibration graph   The slope of this line is very close to one, which indicates that the activities measured by both methods are almost identical. Also the value of coefficient of determination (R2) indicated that the data points fit very well to the trend line.              y = 0.9974x + 0.0024R² = 0.98790.10.150.20.250.30.350.1 0.15 0.2 0.25 0.3 0.35Activity measured by ion chamberActivity estimated by calibration graph  64Chapter  4  Conclusion  The position and strengths of individual brachytherapy seeds loaded into implant needles were verified within to 5% uncertainty using a flat panel detector. Based on the correlation results, the mean reading for a large number of seeds can be obtained with less than 1% uncertainty. Although some improvement in the accuracy might be achievable by more thorough characterization of the detector and its response, the magnitude of the uncertainties relative to those expected on the basis of distance and attenuation effects caused by needle construction and dimensions suggests that they are more likely due to these factors. The uncertainties obtained are adequate for quality assurance of preloaded brachytherapy needles, since, given that all the seeds have already been assayed by the supplier prior to needle loading,  the purpose of the image detector based system is to detect gross errors in the shipment.  Future development of this work would include design of a new autoclavable jig that accommodates as many needles as possible over the entire active area of the detector, so that the quality assurance tests can be performed as efficiently as possible.  Also, a motorized system could be developed to shift the needles along the grooves of the jig to find the maximum reading, instead of doing it manually. Automation of this step would increase the speed and precision of the measurements and would enable more efficient verification of needles loaded with seeds at uneven spacings that do not match the openings in the jig.    Results showed a good correlation between detector reading and radioactive source strength of up to 6 individual seeds loaded into needles as trains with 1 cm or greater spacing between seeds. In addition, the detector has a sufficiently large image capture area to allow several needles (10-15) to be analyzed simultaneously. Good correlation between detector reading and exposure time, as well as between detector reading and source strength, demonstrated the linear characteristic of the detector. By analyzing the images of the seeds, it is possible not only to detect miscalibrated seeds in the needle, but also to verify that seeds are in the correct positions in the source train.  In conclusion, this study showed that a flat panel detector dosimetry system allowed simultaneous characterization in a sterile environment of 100% of seeds used in a brachytherapy treatment procedure. The application of this method is not limited to prostate   65brachytherapy and could be used for other interstitial techniques, such as breast brachytherapy, as well.                                66References 1. Canadian Cancer Society's Advisory Committee on Cancer Statistics, Canadian Public Policy Collection, Canadian Health Research Collection. Canadian Cancer Statistics 2013: Special Topic: Liver Cancer. Canadian Cancer Society; 2013.  2. Greene KL, Zietman A, Carroll P, et al. Prostate Specific Antigen Best Practice Statement: 2009 Update. J Urol. 2009;182:2232-2241.  3. Humphrey PA. Gleason grading and prognostic factors in carcinoma of the prostate. Modern Pathology. 2004;17:292-306.  4. Perlmutter MA, Lepor H. Androgen deprivation therapy in the treatment of advanced prostate cancer. Reviews in urology. 2007;9 Suppl 1:S3-S8.  5. Maddock PG. Brachytherapy sources and applicators. Semin Oncol Nurs. 1987; 3:15-22.  6. Fichtinger G. MO-B-203-01: Robotic Brachytherapy - Overview of Robotic Brachytherapy Approaches and Synergistic Applications. Med Phys. 2010;37:3336.  7. Holm HH. The history of interstitial brachytherapy of prostatic cancer. Semin Surg Oncol. 1997;13:431-437.  8. Aronowitz JN. Don Lawrence and the "k-capture" revolution. Brachytherapy. 2010;9:373-381.  9. http://physics.nist.gov/PhysRefData/Xcom/html/xcom1.html. April 9th 2014.  10. Attix FH. Introduction to Radiological Physics and Radiation Dosimetry. New York: Wiley; 1986.  11. Dose for Nonelectronic Equilibrium Conditions. Radiat Res. 1958;9:399-410.  12. Nath R, Yue N, Liu L. On the depth of penetration of photons and electrons for intravascular brachytherapy. Cardiovasc Radiat Med. 1999;1:72-79.  13. Burlin TE. A general theory of cavity ionisation. Br J Radiol. 1966;39:727.  14. Rivard MJ, Coursey BM, DeWerd LA, et al. Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations. Med Phys. 2004;31:633.  15. Ravinder Nath, Lowell L. Anderson, Gary Luxton, Keith A. Weaver, Jeffrey F. Williamson, Ali S. Meigooni. Dosimetry of interstitial brachytherapy sources:   67Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43. Med Phys. 1995;22:209.  16. American Association of Physicists in Medicine. Supplement to the 2004 update of the AAPM Task Group No. 43 Report. Med Phys. 2007;34:2187-2205.  17. A Handbook of Radioactivity Measurements Procedures, 2nd Edition: (Report no. 58). National Council on Radiation Protection and Measurements (NCRP); 1985.  18. Nag S, Zelefsky MJ, Ciezki JP, et al. Intraoperative planning and evaluation of permanent prostate brachytherapy: report of the American Brachytherapy Society. Int J Radiat Oncol Biol Phys. 2001;51:1422.  19. Prostate brachytherapy postimplant dosimetry: Automatic plan reconstruction of stranded implants. Med Phys. 2010;38:327-342.  20. Butler WM, Merrick GS, Lief JH, Dorsey AT. Comparison of seed loading approaches in prostate brachytherapy. Med Phys. 2000;27:381.  21. Ravinder Nath, Lowell L. Anderson, Jerome A. Meli, Arthur J. Olch, Judith Anne Stitt, Jeffrey F. Williamson. Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56. Med Phys. 1997;24:1557.  22. Fuller DB, Koziol JA, Feng AC. Prostate brachytherapy seed migration and dosimetry: analysis of stranded sources and other potential predictive factors. Brachytherapy. 2004;3:10-19.  23. A dosimetric uncertainty analysis for photon-emitting brachytherapy sources: Report of AAPM Task Group No. 138 and GEC-ESTRO. Med Phys. 2011;38:782-801.  24. Gerald J. Kutcher, Goran K. Svensson, Mona Weller, et al. Comprehensive QA for radiation oncology: Report of AAPM Radiation Therapy Committee Task Group 40. Med Phys. 1994;21:581.  25. Wayne M. Butler, Bruce R. Thomadsen, William S. Bice J, et al. Third-party brachytherapy source calibrations and physicist responsibilities: Report of the AAPM Low Energy Brachytherapy Source Calibration Working Group. Med Phys. 2008;35:3860.  26. Yan Yu, Lowell L. Anderson, Zuofeng Li, et al. Permanent prostate seed implant brachytherapy: Report of the American Association of Physicists in Medicine Task Group No. 64. Med Phys. 1999;26:2054.  27. Yue NJ, Haffty BG, Yue J. On the assay of brachytherapy sources. Med Phys. 2007;34:1975-1982.    6828. Butler WM, Williamson JF, Yue NJ, et al. Third party brachytherapy seed calibrations and physicist responsibilities. Med Phys. 2006;33:247.  29. Wan S, Joshi CP, Carnes G, Schreiner LJ. Evaluation of an automated seed loader for seed calibration in prostate brachytherapy. Journal of applied clinical medical physics / American College of Medical Physics. 2006;7:115-125.  30. Ramos LI, Martinez Monge R. Sampling size in the verification of manufactured-supplied air kerma strengths. Med Phys. 2005;32:3375.  31. Douglas P. Rosenzweig, Michael C. Schell, Yan Yu. Toward a statistically relevant calibration end point for prostate seed implants. Med Phys. 2000;27:144.  32. http://www.seedos.co.uk/SeeDOS_Ltd-ivb_1000brochuremain1159-02.pdf. April 10th  2014 33. Larry A. DeWerd, John A. Micka, Shannon M. Holmes, Tim D. Bohm. Calibration of multiple LDR brachytherapy sources. Med Phys. 2006;33:3804.  34. Butler WM, Dorsey AT, Nelson KR, Merrick GS. Quality Assurance Calibration of 125I Rapid Strand in a Sterile Environment. Int J Radiat Oncol Biol Phys. 1998;41:217-222.  35. Mellenberg J,D E., Pennington EC. 103Pd loaded cartridge air kerma strength verification. Medical dosimetry : official journal of the American Association of Medical Dosimetrists. 1999;24:73-75.  36. Furutani S, Saze T, Ikushima H, et al. Quality assurance of I-125 seeds for prostate brachytherapy using an imaging plate. Int J Radiat Oncol Biol Phys. 2006;66:603-609.  37. Larry A. DeWerd, John A. Micka, Shannon M. Holmes, Tim D. Bohm. Calibration of multiple LDR brachytherapy sources. Med Phys. 2006;33:3804.  38. Heintz BH, Wallace RE, Hevezi JM. Comparison of I-125 sources used for permanent interstitial implants. Med Phys. 2001;28:671.  39. Chiu-Tsao S, Medich D, Munro 3, John. The use of new GAFCHROMIC EBT film for 125I seed dosimetry in Solid Water phantom. Med Phys. 2008;35:3787.  40. Symonds-Tayler JRN, Partridge M, Evans PM. An electronic portal imaging device for transit dosimetry. Phys Med Biol. 1997;42:2273.  41. Spahn M. Flat detectors and their clinical applications. Eur Radiol. 2005;15:1934-1947.  42. Cowen AR, Davies AG, Sivananthan MU. The design and imaging characteristics of dynamic, solid-state, flat-panel x-ray image detectors for digital fluoroscopy and fluorography. Clin Radiol. 2008;63:1073-1085.    6943. Konstantinidis AC, Szafraniec MB, Speller RD, Olivo A. The Dexela 2923 CMOS X-ray detector: A flat panel detector based on CMOS active pixel sensors for medical imaging applications. Nuclear Instruments and Methods in Physics Research, A. 2012;689:12.   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0167373/manifest

Comment

Related Items