Table of contents

The UK, European and IAEA protocols for breast dosimetry in mammography use tabulations of conversion factors, which relate measurements of incident air kerma to the mean glandular dose to the breast. To supplement the existing tabulations, a Monte Carlo computer program has been used to calculate conversion factors for the high-energy spectra used for contrast enhanced digital mammography. The calculations were made for the x-ray spectra from a tungsten target (tube voltage range 40–50 kV) filtered by 0.28, 0.30 and 0.32 mm of copper, and from molybdenum and rhodium targets (tube voltage range 40–49 kV), each filtered by 0.30 mm of copper. The g-factors for all of these spectra were plotted for each breast thickness as a function of half value layer (HVL) and were found to lie on smooth curves within 0.3%. These reflect the fact that the characteristic x-rays present in the spectra from molybdenum and rhodium are heavily filtered and all the spectra are essentially Bremsstrahlung. As a consequence, the s-factor previously used in the dosimetry protocols to adjust for different target/filter combinations can be taken as unity for all of the spectra considered. Tables of g-factors and c-factors are provided for breast thicknesses in the range 20–110 mm and HVLs in the range 2.4–3.6 mm of aluminium. The tables of c-factors are given for breast glandularities in the range 0.1%–100% and for typical glandularities for women in the age bands 40–49 and 50–64 attending the UK national breast screening programme.

Clinical studies have established a strong benefit from adjuvant mild hyperthermia (HT) to radio- and chemotherapy for many tumor sites, including the head and neck (H&N). The recently developed HYPERcollar allows the application of local radiofrequency HT to tumors in the entire H&N. Treatment quality is optimized using electromagnetic and thermal simulators and, whenever placement risk is tolerable, assessed using invasively placed thermometers. To replace the current invasive procedure, we are investigating whether magnetic resonance (MR) thermometry can be exploited for continuous and 3D thermal dose assessment. In this work, we used our simulation tools to design an MR compatible laboratory prototype applicator. By simulations and measurements, we showed that the redesigned patch antennas are well matched to 50 Ω (S11<−10 dB). Simulations also show that, using 300 W input power, a maximum specific absorption rate (SAR) of 100 W kg⁻¹ and a temperature increase of 4.5 °C in 6 min is feasible at the center of a cylindrical fat/muscle phantom. Temperature measurements using the MR scanner confirmed the focused heating capabilities and MR compatibility of the setup. We conclude that the laboratory applicator provides the possibility for experimental assessment of the feasibility of hybrid MR-HT in the H&N region. This versatile design allows rigorous analysis of MR thermometry accuracy in increasingly complex phantoms that mimic patients' anatomies and thermodynamic characteristics.

Recent studies emphasized the role of the bone lacuno-canalicular network (LCN) in the understanding of bone diseases such as osteoporosis. However, suitable methods to investigate this structure are lacking. The aim of this paper is to introduce a methodology to segment the LCN from three-dimensional (3D) synchrotron radiation nano-CT images. Segmentation of such structures is challenging due to several factors such as limited contrast and signal-to-noise ratio, partial volume effects and huge number of data that needs to be processed, which restrains user interaction. We use an approach based on minimum-cost paths and geodesic voting, for which we propose a fully automatic initialization scheme based on a tessellation of the image domain. The centroids of pre-segmented lacunæ are used as Voronoi-tessellation seeds and as start-points of a fast-marching front propagation, whereas the end-points are distributed in the vicinity of each Voronoi-region boundary. This initialization scheme was devised to cope with complex biological structures involving cells interconnected by multiple thread-like, branching processes, while the seminal geodesic-voting method only copes with tree-like structures. Our method has been assessed quantitatively on phantom data and qualitatively on real datasets, demonstrating its feasibility. To the best of our knowledge, presented 3D renderings of lacunæ interconnected by their canaliculi were achieved for the first time.

Specific absorbed fractions (SAFs) have been calculated as a function of the content in the urinary bladder in order to allow more realistic calculations of the absorbed dose to the bladder wall. The SAFs were calculated using the urinary bladder anatomy from the ICRP male and female adult reference computational phantoms. The urinary bladder and its content were approximated by a sphere with a wall of constant mass, where the thickness of the wall depended on the amount of urine in the bladder. SAFs were calculated for males and females with 17 different urinary bladder volumes from 10 to 800 mL, using the Monte Carlo computer program MCNP5, at 25 energies of mono-energetic photons and electrons ranging from 10 KeV to 10 MeV. The decay was assumed to be homogeneously distributed in the urinary bladder content and the urinary bladder wall, and the mean absorbed dose to the urinary bladder wall was calculated. The Monte Carlo simulations were validated against measurements made with thermoluminescent dosimeters. The SAFs obtained for a urine volume of 200 mL were compared to the values calculated for the urinary bladder wall using the adult reference computational phantoms. The mean absorbed dose to the urinary wall from ¹⁸F-FDG was found to be 77 µGy/MBq formales and 86 µGy/MBq for females, while for ^99mTc-DTPA the mean absorbed doses were 80 µGy/MBq for males and 86 µGy/MBq for females. Compared to calculations using a constant value of the SAF from the adult reference computational phantoms, the mean absorbed doses to the bladder wall were 60% higher for ¹⁸F-FDG and 30% higher for ^99mTc-DTPA using the new SAFs.

The GATE Monte Carlo simulation platform based on the Geant4 toolkit is under constant improvement for dosimetric calculations. In this study, we explore its use for the dosimetry of the preclinical targeted radiotherapy of melanoma using a new specific melanin-targeting radiotracer labeled with iodine 131. Calculated absorbed fractions and S values for spheres and murine models (digital and CT-scan-based mouse phantoms) are compared between GATE and EGSnrc Monte Carlo codes considering monoenergetic electrons and the detailed energy spectrum of iodine 131. The behavior of Geant4 standard and low energy models is also tested. Following the different authors' guidelines concerning the parameterization of electron physics models, this study demonstrates an agreement of 1.2% and 1.5% with EGSnrc, respectively, for the calculation of S values for small spheres and mouse phantoms. S values calculated with GATE are then used to compute the dose distribution in organs of interest using the activity distribution in mouse phantoms. This study gives the dosimetric data required for the translation of the new treatment to the clinic.

Recent developments have shown that high resolution phase contrast x-ray computed tomography (CT) of the breast can be performed at clinically compatible doses. Results have yet been obtained in vitro on full breasts, and the clinical translation of the technique seems more and more possible. This work presents a method to quickly estimate the average dose in the organ using the software GATE. The influence of different parameters on the dose distribution, like breast composition and thickness, and for preclinical test, the presence of a skin/PMMA external layer, has been investigated. Several correction factors, to be applied to the given dose database, are also introduced to allow the use of these results in geometries different from those studied here. An energy optimization study is presented that considers also the influence on the energy choice of x-ray detector. A simple analytical method to estimate the best energy that minimizes the dose-transmittance ratio in CT imaging is presented and compared with the results of simulations.

Tumor motion caused by respiration is an important issue in image-guided radiotherapy. A 2D/4D matching method between 4D volumes derived from cone beam computed tomography (CBCT) and 2D fluoroscopic images was implemented to track the tumor motion without the use of implanted markers. In this method, firstly, 3DCBCT and phase-rebinned 4DCBCT are reconstructed from cone beam acquisition. Secondly, 4DCBCT volumes and a streak-free 3DCBCT volume are combined to improve the image quality of the digitally reconstructed radiographs (DRRs). Finally, the 2D/4D matching problem is converted into a 2D/2D matching between incoming projections and DRR images from each phase of the 4DCBCT. The diaphragm is used as a target surrogate for matching instead of using the tumor position directly. This relies on the assumption that if a patient has the same breathing phase and diaphragm position as the reference 4DCBCT, then the tumor position is the same. From the matching results, the phase information, diaphragm position and tumor position at the time of each incoming projection acquisition can be derived. The accuracy of this method was verified using 16 candidate datasets, representing lung and liver applications and one-minute and two-minute acquisitions. The criteria for the eligibility of datasets were described: 11 eligible datasets were selected to verify the accuracy of diaphragm tracking, and one eligible dataset was chosen to verify the accuracy of tumor tracking. The diaphragm matching accuracy was 1.88 ± 1.35 mm in the isocenter plane and the 2D tumor tracking accuracy was 2.13 ± 1.26 mm in the isocenter plane. These features make this method feasible for real-time marker-free tumor motion tracking purposes.

We present the development and application of a phantom for assessment and optimization of fat suppression over a large field-of-view in diffusion-weighted magnetic resonance imaging at 1.5 T and 3 T. A Perspex cylinder (inner diameter 185 mm, height 300 mm) which contains a second cylinder (inner diameter 140 mm) was constructed. The inner cylinder was filled with water doped with copper sulphate and sodium chloride and the annulus was filled with corn oil, which closely matches the spectrum and longitudinal relaxation times of subcutaneous abdominal fat. Placement of the phantom on the couch at 45° to the z-axis presented an elliptical cross-section, which was of a similar size and shape to axial abdominal images. The use of a phantom for optimization of fat suppression allowed quantitative comparison between studies without the differences introduced by variability between human subjects. We have demonstrated that the phantom is suitable for selection of inversion delay times, spectral adiabatic inversion recovery delays and assessment of combinatorial methods of fat suppression. The phantom is valuable in protocol development and the assessment of new techniques, particularly in multi-centre trials.

Due to the high atomic number of gold nanoparticles (GNPs), they are known as new radiosensitizer agents for enhancing the efficiency of superficial radiotherapy techniques by increasing the dose absorbed in tumor cells wherein they can be accumulated selectively. The aim of this study was to compare the effect of various common low energy levels of orthovoltage x-rays and megavoltage γ-rays (Co-60) on enhancing the therapeutic efficiency of HeLa cancer cells in the presence of conjugated folate and non-conjugated (pegylated) GNPs. To achieve this, GNPs with an average diameter of 52 nm were synthesized and conjugated to folic acid molecules. Pegylated GNPs with an average diameter of 47 nm were also synthesized and used as non-conjugated folate GNPs. Cytotoxicity assay of the synthesized folate-conjugated and pegylated GNPs was performed using different levels of nanoparticle concentration incubated with HeLa cells for 24 h. The radiosensitizing effect of both the conjugated and pegylated GNPs on the cells at a concentration of 50 µM was compared using MTT as well as clonogenic assays after exposing them to 2 Gy ionizing radiation produced by an orthovoltage x-ray machine at four different kVps and γ-rays of a Co-60 unit. Significant differences were noted among various irradiated groups with and without the folate conjugation, with an average dose enhancement factor (DEF) of 1.64 ± 0.05 and 1.35 ± 0.05 for the folate-conjugated and pegylated GNPs, respectively. The maximum DEF was obtained with the 180 kVp x-ray beam for both of the GNPs. Folate-conjugated GNPs can significantly enhance the cell killing potential of orthovoltage x-ray energies (especially at 180 kVp) in folate receptor-expressing cancer cells, such as HeLa, in superficial radiotherapy techniques.

Today, quantitative analysis of three-dimensional (3D) dynamics of the left ventricle (LV) cannot be performed directly in the catheter lab using a current angiographic C-arm system, which is the workhorse imaging modality for cardiac interventions. Therefore, myocardial wall analysis is completely based on the 2D angiographic images or pre-interventional 3D/4D imaging. In this paper, we present a complete framework to study the ventricular wall motion in 4D (3D+t) directly in the catheter lab. From the acquired 2D projection images, a dynamic 3D surface model of the LV is generated, which is then used to detect ventricular dyssynchrony. Different quantitative features to evaluate LV dynamics known from other modalities (ultrasound, magnetic resonance imaging) are transferred to the C-arm CT data. We use the ejection fraction, the systolic dyssynchrony index a 3D fractional shortening and the phase to maximal contraction (ϕ_{i, max}) to determine an indicator of LV dyssynchrony and to discriminate regionally pathological from normal myocardium. The proposed analysis tool was evaluated on simulated phantom LV data with and without pathological wall dysfunctions. The LV data used is publicly available online at https://conrad.stanford.edu/data/heart. In addition, the presented framework was tested on eight clinical patient data sets. The first clinical results demonstrate promising performance of the proposed analysis tool and encourage the application of the presented framework to a larger study in clinical practice.

The present state of modeling radio-induced effects at the cellular level does not account for the microscopic inhomogeneity of the nucleus from the non-aqueous contents (i.e. proteins, DNA) by approximating the entire cellular nucleus as a homogenous medium of water. Charged particle track-structure calculations utilizing this approximation are therefore neglecting to account for approximately 30% of the molecular variation within the nucleus. To truly understand what happens when biological matter is irradiated, charged particle track-structure calculations need detailed knowledge of the secondary electron cascade, resulting from interactions with not only the primary biological component—water-–but also the non-aqueous contents, down to very low energies. This paper presents our work on a generic approach for calculating low-energy interaction cross-sections between incident charged particles and individual molecules. The purpose of our work is to develop a self-consistent computational method for predicting molecule-specific interaction cross-sections, such as the component molecules of DNA and proteins (i.e. nucleotides and amino acids), in the very low-energy regime. These results would then be applied in a track-structure code and thereby reduce the homogenous water approximation. The present methodology—inspired by seeking a combination of the accuracy of quantum mechanics and the scalability, robustness, and flexibility of Monte Carlo methods—begins with the calculation of a solution to the many-body Schrödinger equation and proceeds to use Monte Carlo methods to calculate the perturbations in the internal electron field to determine the interaction processes, such as ionization and excitation. As a test of our model, the approach is applied to a water molecule in the same method as it would be applied to a nucleotide or amino acid and compared with the low-energy cross-sections from the GEANT4-DNA physics package of the Geant4 simulation toolkit for the energy ranges of 7 eV to 1 keV.

The purpose of this study was to examine the use of three physical image quality metrics in the calibration of an automatic exposure control (AEC) device for chest radiography with a computed radiography (CR) imaging system. The metrics assessed were signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR) and mean effective noise equivalent quanta (eNEQ_m), all measured using a uniform chest phantom. Subsequent calibration curves were derived to ensure each metric was held constant across the tube voltage range. Each curve was assessed for its clinical appropriateness by generating computer simulated chest images with correct detector air kermas for each tube voltage, and grading these against reference images which were reconstructed at detector air kermas correct for the constant detector dose indicator (DDI) curve currently programmed into the AEC device. All simulated chest images contained clinically realistic projected anatomy and anatomical noise and were scored by experienced image evaluators. Constant DDI and CNR curves do not appear to provide optimized performance across the diagnostic energy range. Conversely, constant eNEQ_m and SNR do appear to provide optimized performance, with the latter being the preferred calibration metric given as it is easier to measure in practice. Medical physicists may use the SNR image quality metric described here when setting up and optimizing AEC devices for chest radiography CR systems with a degree of confidence that resulting clinical image quality will be adequate for the required clinical task. However, this must be done with close cooperation of expert image evaluators, to ensure appropriate levels of detector air kerma.

Recent studies have suggested that the characteristics of prompt gammas (PGs) emitted from excited nuclei during proton therapy are advantageous for determining beam range during treatment delivery. Since PGs are only emitted while the beam is on, the feasibility of using PGs for online treatment verification depends greatly on the design of highly efficient detectors. The purpose of this work is to characterize how PG detection changes as a function of distance from the patient as a means of guiding the design and usage of clinical PG imaging detectors. Using a Monte Carlo model (GEANT4.9.4) we studied the detection rate (PGs per incident proton) of a high purity germanium detector for both the total PG emission and the characteristic 6.13 MeV PG emission from ¹⁶O emitted during proton irradiation. The PG detection rate was calculated as a function of distance from the isocenter of the proton treatment nozzle for: (1) a water phantom irradiated with a proton pencil beam and (2) a prostate patient irradiated with a scanning beam proton therapy treatment field (lateral field size: ∼6 cm × 6 cm, beam range: 23.5 cm). An analytical expression of the PG detection rate as a function of distance from isocenter, detector size, and proton beam energy was then developed. The detection rates were found to be 1.3 × 10⁻⁶ for oxygen and 3.9 × 10⁻⁴ for the total PG emission, respectively, with the detector placed 11 cm from isocenter for a 40 MeV pencil beam irradiating a water phantom. The total PG detection rate increased by ∼85 ± 3% for beam energies greater than 150 MeV. The detection rate was found to be approximately 2.1 × 10⁻⁶ and 1.7 × 10⁻³ for oxygen and total PG emission, respectively, during delivery of a single pencil beam during a scanning beam treatment for prostate cancer. The PG detection rate as a function of distance from isocenter during irradiation of a water phantom with a single proton pencil beam was described well by the model of a point source irradiating a cylindrical detector of a known diameter over the range of beam energies commonly used for proton therapy. For the patient studies, it was necessary to divide the point source equation by an exponential factor in order to correctly predict the falloff of the PG detection rate as a function of distance from isocenter.

A promising, new, in vivo prostate dosimetry system has been developed for clinical radiation therapy. This work outlines the preliminary end-to-end testing of the accuracy and precision of the new OARtrac scintillation dosimetry system. We tested 94 calibrated plastic scintillation detector (PSD) probes before their final integration into endorectal balloon assemblies. These probes had been calibrated at The University of Texas MD Anderson Cancer Center Dosimetry Laboratory. We used a complete clinical OARtrac system including the PSD probes, charge coupled device camera monitoring system, and the manufacturer's integrated software package. The PSD probes were irradiated at 6 MV in a Solid Water® phantom. Irradiations were performed with a 6 MV linear accelerator using anterior–posterior/posterior–anterior matched fields to a maximum dose of 200 cGy in a 100 cm source-axis distance geometry. As a whole, the OARtrac system has good accuracy with a mean error of 0.01% and an error spread of ±5.4% at the 95% confidence interval. These results reflect the PSD probes' accuracy before their final insertion into endorectal balloons. Future work will test the dosimetric effects of mounting the PSD probes within the endorectal balloon assemblies.

The objective of this study was to demonstrate the potential benefits of using high energy x-rays for phase sensitive breast imaging through a comparison with conventional mammography imaging. We compared images of a contrast-detail phantom acquired on a prototype phase sensitive x-ray imaging system with images acquired on a commercial flat panel digital mammography unit. The phase contrast images were acquired using a micro-focus x-ray source with a 50 µm focal spot at 120 kVp and 4.5 mAs, with a magnification factor of 2.46 and a 50 µm pixel pitch. A phase attenuation duality-based phase retrieval algorithm that requires only a single phase contrast image was applied. Conventional digital mammography images were acquired at 27 kVp, 131 mAs and 28 kVp, 54 mAs. For the same radiation dose, both the observer study and signal-to-noise ratio (SNR)/figure of merit comparisons indicated a large improvement by the phase retrieved image as compared to the clinical system for the larger disc sizes, but the improvement was not enough to detect the smallest discs. Compared to the double dose image acquired with the clinical system, the observer study also indicated that the phase retrieved image provided improved detection capabilities for all disc sizes except the smallest discs. Thus the SNR improvement provided by phase contrast imaging is not yet enough to offset the noise reduction provided by the clinical system at the doubled dose level. However, the potential demonstrated by this study for high energy phase sensitive x-ray imaging to improve lesion detection and reduce radiation dose in mammography warrants further investigation of this technique.

This work investigated the differences between multileaf collimator (MLC) positioning accuracy determined using either log files or electronic portal imaging devices (EPID) and then assessed the possibility of reducing patient specific quality control (QC) via phantom-less methodologies. In-house software was developed, and validated, to track MLC positional accuracy with the rotational and static gantry picket fence tests using an integrated electronic portal image. This software was used to monitor MLC daily performance over a 1 year period for two Varian TrueBeam linear accelerators, with the results directly compared with MLC positions determined using leaf trajectory log files. This software was validated by introducing known shifts and collimator errors. Skewness of the MLCs was found to be 0.03 ± 0.06° (mean ±1 standard deviation (SD)) and was dependent on whether the collimator was rotated manually or automatically. Trajectory log files, analysed using in-house software, showed average MLC positioning errors with a magnitude of 0.004 ± 0.003 mm (rotational) and 0.004 ± 0.011 mm (static) across two TrueBeam units over 1 year (mean ±1 SD). These ranges, as indicated by the SD, were lower than the related average MLC positioning errors of 0.000 ± 0.025 mm (rotational) and 0.000 ± 0.039 mm (static) that were obtained using the in-house EPID based software. The range of EPID measured MLC positional errors was larger due to the inherent uncertainties of the procedure. Over the duration of the study, multiple MLC positional errors were detected using the EPID based software but these same errors were not detected using the trajectory log files. This work shows the importance of increasing linac specific QC when phantom-less methodologies, such as the use of log files, are used to reduce patient specific QC. Tolerances of 0.25 mm have been created for the MLC positional errors using the EPID-based automated picket fence test. The software allows diagnosis of any specific leaf that needs repair and gives an indication as to the course of action that is required.