Table of contents

Intrafraction motion caused by breathing requires increased treatment margins for chest and abdominal radiotherapy and may lead to `motion artefacts' in dose distributions during intensity modulated radiotherapy (IMRT). Technologies such as gated radiotherapy may significantly increase the treatment time, while breath-hold techniques may be poorly tolerated by pulmonarily compromised patients. A solution that allows reduced margins and dose distribution artefacts, without compromising delivery time, is to synchronously follow the target motion by adapting the x-ray beam using a dynamic multileaf collimator (MLC), i.e. motion adaptive x-ray therapy, or MAX-T for short. Though the target is moving with time, in the MAX-T beam view the target is static. The MAX-T method superimposes the target motion due to respiration onto the beam originally planned for delivery. Thus during beam delivery the beam is dynamically changing position with respect to the isocentre using a dynamic MLC, the leaf positions of which are dependent upon the target position. Synchronization of the MLC motion and target motion occurs using respiration gated radiotherapy equipment. The concept and feasibility of MAX-T and the capability of the treatment machine to deliver such a treatment were investigated by performing measurements for uniform and IMRT fields using a mechanical sinusoidal oscillator to simulate target motion. Target dose measurements obtained using MAX-T for a moving target were found to be equivalent to those delivered to a static target by a static beam.

A detailed tomotherapy inverse treatment planning method is described which incorporates leakage and head scatter corrections during each iteration of the optimization process, allowing these effects to be directly accounted for in the optimized dose distribution. It is shown that the conventional inverse planning method for optimizing incident intensity can be extended to include a `concurrent' leaf sequencing operation from which the leakage and head scatter corrections are determined. The method is demonstrated using the steepest-descent optimization technique with constant step size and a least-squared error objective. The method was implemented using the MATLAB scientific programming environment and its feasibility demonstrated for 2D test cases simulating treatment delivery using a single coplanar rotation. The results indicate that this modification does not significantly affect convergence of the intensity optimization method when exposure times of individual leaves are stratified to a large number of levels (>100) during leaf sequencing. In general, the addition of aperture dependent corrections, especially `head scatter', reduces incident fluence in local regions of the modulated fan beam, resulting in increased exposure times for individual collimator leaves. These local variations can result in 5% or greater local variation in the optimized dose distribution compared to the uncorrected case. The overall efficiency of the modified intensity optimization algorithm is comparable to that of the original unmodified case.

The aims of this study were to determine whether the location on the tibia measured by ¹⁰⁹Cd-based K-shell x-ray fluorescence (XRF) affected the measurement result and its uncertainty, and whether higher tibia lead levels at the extremities of the tibia and/or inhomogeneity in the distribution of lead in the tibia could be inferred therefrom.

Replicate XRF measurements were performed at multiple locations on ten adult cadaver intact legs and on nine bare tibiæ dissected from them.

Mean lead levels in the bare tibiæ ranged from 16 to 48 µg Pb per g of bone mineral. Bare tibia measurements showed that both the XRF result and its uncertainty increased towards the proximal and distal ends of the tibia. The XRF result decreased away from the medial-lateral mid-point of the tibia, but XRF uncertainty was not significantly affected. Intact leg measurements showed no effect of proximal-distal location on XRF result but did show an effect on XRF uncertainty.

We conclude that the XRF method used can determine the differences in bone lead level resulting from the more trabecular composition at the ends of the tibia, and we present limited evidence for localized regions of low tibia lead level.

The frequency-domain multiple-distance (FDMD) method is capable of measuring the absolute absorption and reduced scattering coefficients of optically turbid media. Absolute measurement of absorption at two near-infrared (NIR) wavelengths makes possible the quantitation of tissue haemoglobin concentration and tissue haemoglobin oxygen-saturation (StO₂). However, errors are introduced by the uncertainties of background absorption and the dissimilarities between real tissues and the simplified mathematical model on which these measurements are based. An FDMD-based tissue instrument has been used for the monitoring of tissue haemoglobin concentration and oxygenation in the brain of newborn piglets during periods of hypoxia and hyperoxia. These tissue haemoglobin saturation values were compared with arterial saturation (SaO₂) and venous saturation (SvO₂) measured by blood gas analyses. A linear correlation was observed between StO₂ and the average of SaO₂ and SvO₂. However, StO₂ is not equal to any fixed weighted average of SaO₂ and SvO₂ unless we introduce an effective background tissue absorption. The magnitude of the background absorption was about 0.08 cm^-1 at 758 nm and 0.06 cm^-1 at 830 nm, and it was nearly consistent between piglets. The origin of this `effective' background absorption may be real, an artefact caused by the application of a simplified model to a complex sample, or a combination of factors.

Investigating the effect of low-dose radiation exposure on cells using assays of colony-forming ability requires large cell samples to maintain statistical accuracy. Manually counting the resulting colonies is a laborious task in which consistent objectivity is hard to achieve. This is true especially with some mammalian cell lines which form poorly defined or `fuzzy' colonies, typified by glioma or fibroblast cell lines. A computer-vision-based automated colony counter is presented in this paper. It utilizes novel imaging and image-processing methods involving a modified form of the Hough transform. The automated counter is able to identify less-discrete cell colonies typical of these cell lines. The results of automated colony counting are compared with those from four manual (human) colony counts for the cell lines HT29, A172, U118 and IN1265. The results from the automated counts fall well within the distribution of the manual counts for all four cell lines with respect to surviving fraction (SF) versus dose curves, SF values at 2 Gy (SF₂) and total area under the SF curve (Dbar). From the variation in the counts, it is shown that the automated counts are generally more consistent than the manual counts.

We used a real-skull phantom head to investigate the performances of representative methods for EEG source localization when considering various head models.

We describe several experiments using a montage with current sources located at multiple positions and orientations inside a human skull filled with a conductive medium. The robustness of selected methods based on distributed source models is evaluated as various solutions to the forward problem (from the sphere to the finite element method) are considered.

Experimental results indicate that inverse methods using appropriate cortex-based source models are almost always able to locate the active source with excellent precision, with little or no spurious activity in close or distant regions, even when two sources are simultaneously active.

Superior regularization schemes for solving the inverse problem can dramatically help the estimation of sparse and focal active zones, despite significant approximation of the head geometry and the conductivity properties of the head tissues. Realistic head models are necessary, though, to fit the data with a reasonable level of residual variance.

A prototype x-ray needle, which emits 62.5 kVp x-rays at the tip of a 20 cm long, 4 mm diameter steel needle, has been developed by Titan Pulse Sciences Incorporated (PSI) (Albuquerque, NM) and was tested for its suitability in brachytherapy applications in comparison with a similar device by the Photoelectron Corporation. The depth dose profiles were also compared with those of two common brachytherapy sources (¹²⁵I and ¹⁹²Ir). The depth dose characteristics of the radiation were comparable with the two brachytherapy sources with a slightly reduced attenuation gradient. The dose rate from the x-ray needle tip was relatively isotropic at the needle tip and was continuously adjustable over the range of 0 cGy min^-1 to upwards of 62 cGy min^-1 at a reference distance of 1 cm in air. We detected a significant proportion of x-rays generated along the needle shaft, and not at the needle tip, as intended. The energy spectrum emitted from this device had a peak intensity at 21 keV and an average energy of 28 keV. The beam was attenuated in both aluminium (the first half-value layer being less than 0.1 mm) and in water (50% dose at approximately 2 mm). These studies confirm that although there is potential for a system similar to this one for clinical applications, the simplistic electron guidance used in this particular prototype device limits it to research applications. Further optimization is required in focusing and steering the electron beam to the target, improving x-ray production efficiency and using x-ray target cooling to achieve higher dose rates.

Multiple local minima exist in almost every coplanar or non-coplanar radiotherapy treatment planning problem. We used a global optimization method based on topographical information about the distribution of local minima to find all local minima and select the best as the global minimum. We uniformly select N random points from search regions and construct a topographical graph, from which M (M << N) starting points are selected to launch a local search. Because every seed point was at or near the local minimum, the solutions found by the local search could be used as the final optimization results. We verified this algorithm by applying it to three different clinical cases and comparing the results with those obtained by a local optimization method (sequential quadratic programming). The results show that this algorithm is feasible and efficient.

We have proposed the utilization of `hyper-thermal neutrons' for neutron capture therapy (NCT) from the viewpoint of the improvement in the dose distribution in a human body. In order to verify the improved depth-dose distribution due to hyper-thermal neutron incidence, two experiments were carried out using a test-type hyper-thermal neutron generator at a thermal neutron irradiation field in Kyoto University Reactor (KUR), which is actually utilized for NCT clinical irradiation. From the free-in-air experiment for the spectrum-shift characteristics, it was confirmed that the hyper-thermal neutrons of approximately 860 K at maximum could be obtained by the generator. From the phantom experiment, the improvement effect and the controllability for the depth-dose distribution were confirmed. For example, it was found that the relative neutron depth-dose distribution was about 1 cm improved with the 860 K hyper-thermal neutron incidence, compared to the normal thermal neutron incidence.

Anisotropy functions for low energy interstitial brachytherapy sources are examined. Absolute dose rates around ¹⁰³Pd seed model 200 and ¹²⁵I seed models 6702 and 6711 have been estimated by means of the EGS4 Monte Carlo simulation system. The DLC-136/PHOTX cross section library, water molecular form factors, bound Compton scattering and Doppler broadening of the Compton-scattered photon energy were considered in the calculations. Following the formalism developed by the Interstitial Brachytherapy Collaborative Working Group, anisotropy functions, F(r,θ), have been calculated. Our Monte Carlo results were compared against a limited set of measured data selected from the literature and other Monte Carlo results. Binding corrections and phantom material selection have been found to have no influence on the anisotropy function. The accuracy of the geometrical source models used for the Monte Carlo calculations was validated against experimental measurements of in-air relative fluence at 100 cm from the source. More detailed knowledge about the geometrical design of ¹⁰³Pd seed model 200 is needed in order to improve the agreement with experimentally measured in-air fluence. Values for in-air fluence of ¹²⁵I model 6702 are sensitive to source position within the inner seed cylinder. Excellent agreement between calculated and measured in-air fluence is found for ¹²⁵I model 6711. It was observed that using in-air relative fluence at 100 cm from the source to calculate the anisotropy function yields a less anisotropic dose distribution at distances close to the source than full Monte Carlo simulation, in contradiction with experimental data. Our results have estimated statistical uncertainties of 1%-3% at the 1σ level within clinically relevant regions, but contain systematic uncertainties related to the assumed geometrical details.

Fourier transform Raman spectroscopy was undertaken in the study of irradiated polyacrylamide gels (PAGs) used in 3D radiation dosimetry. By employing correlation techniques, monomer and crosslinker consumption were characterized in the spectra as a function of absorbed dose. The consumption of both monomer and crosslinker is monoexponential up to 13 Gy, although the rates of consumption differ for the two molecules. A sensitivity parameter, D₀, in the exponential function has been used to characterize this difference. Up to 13 Gy, D₀(acr) = 12 ± 2 Gy while D₀(bis) = 8.0 ± 0.5 Gy, indicating that bis is consumed at a greater rate than acrylamide and that bis is the limiting factor in the onset of gel saturation, for a gel composition of 6% by weight total monomer (6%T) and where 3% of the total monomer is crosslinker (50%C). Direct evidence of polymer formation was observed in the Raman spectra of irradiated PAG. Polymer formation is monoexponential to a dose of 13 Gy, with a sensitivity parameter of D₀(poly) = 14 ± 2 Gy. This is in good agreement with the consumption rate of acrylamide. The exponential nature of the polymer formation observed here is compared with existing MRI and x-ray CT dose response measurements previously reported to be linear. The results confirm previous studies indicating that Raman spectroscopy provides a direct and useful tool for characterization of irradiated PAG.

The weak absorption of shortwave infrared light by skin tissues between 700 and 1500 nm offers an important window for diagnosis by optical means. The strong scattering of shortwave infrared light by the skin, however, presents a challenge to the modelling of light propagation through the skin and the understanding of skin optics. We have measured the collimated and diffuse transmittance and diffuse reflectance of porcine skin dermis samples within 30 h post-mortem. Monte Carlo simulations have been performed to inversely determine the absorption coefficient, scattering coefficient and anisotropy factor of the dermis samples in the spectral range from 900 to 1500 nm. We further analyse the sensitivity of the values of the parameters to the experimental errors and inverse calculation procedures. The state of the cellular integrity of the skin samples following optical measurements was verified using transmission electron microscopy. These results were correlated to study post-mortem effects on the in vitro optical properties of porcine dermis. We concluded that for samples stored within crushed ice for up to 30 h post-mortem the wavelength dependence of optical properties of the dermis remains unchanged while the values of the parameters vary moderately due to modification of the water content of the tissue.

Present-day regional hyperthermia treatment planning systems are limited to centimetre resolution. To obtain CT-resolution SAR distributions, a method called quasistatic zooming has been developed: using the centimetre-resolution -field distribution and the CT-resolution tomogram, the CT-resolution SAR distribution is obtained. For a low frequency of 10 MHz this method has been validated sucessfully using CT-resolution SAR computations.

It appears that these CT-resolution SAR distributions are completely different from centimetre-resolution SAR distributions, indicating the necessity for high-resolution SAR modelling. Using the presented zooming technique, reliable CT-resolution SAR modelling is now possible with relatively short computation times. So far, the zooming method has only been validated for low frequencies, but clinically relevant frequencies appear to be possible.

We have developed a coaxial measurement system for determining the time and temperature dependence of the dielectric properties of bovine liver at 915 MHz during heating. Our data suggest that changes in dielectric properties due to heating are dominated by the relaxation response of two tissue components: tissue water and proteins. At temperatures above 60 °C, the effects of these two components contribute to increases of up to 100% and 5% in the values of conductivity and permittivity respectively. Changes due to tissue water content were found to be reversible with temperature, while changes due to protein denaturation were found to be permanent. The temperature coefficients for reversible changes were found to be 1.82±0.28%  °C^-1 and -0.130±(5.9×10^-2)%  °C^-1 for conductivity and permittivity respectively. The critical temperatures and activation energies leading to irreversible changes in conductivity and permittivity were determined using Arrhenius analysis. Frequency factors of (1.14±0.27)×10⁴³ s^-1 and (1.95±0.49)×10³⁶ s^-1 were determined for permittivity and conductivity respectively. The activation energies were calculated to be 70.7±15.8 kcal mol^-1 for permittivity and 60.1±14.0 kcal mol^-1 for conductivity.

This paper shows the importance of using a cell model with the proper geometry, orientation and internal structure to study possible cellular effects from direct radiofrequency exposure. For this purpose, the electric field intensity is calculated, using the finite element numerical technique, in single- and multilayer spherical, cylindrical and ellipsoidal mammalian cell models exposed to linearly polarized electromagnetic plane waves of frequencies 900 and 2450 MHz. An extensive analysis is performed on the influence that the cell geometry and orientation with respect to the external field have in the value of the electric field induced in the membrane and cytoplasm. We also show the significant role that the cytoplasmic and extracellular bound water layers play in determining the electric field intensity for the cylindrical and ellipsoidal cell models. Finally, a study of the mutual interactions between cells shows that polarizing effects between cells significantly modify the values of field intensity within the cell.

We present an investigation of the fluoroscopic imaging and dosimetric performances of iodine- and gadolinium-based vascular contrast agents in combination with K-absorption edge filters of atomic numbers between 50 (tin) and 82 (lead). These combinations were studied using a theoretical model for a range of diagnostic x-ray spectra (55 to 100 kVp) and for water phantoms representative of thin and thick anatomies. Performance was characterized by radiographic contrast, a derived image quality index, the patient integral and entrance skin doses, and the x-ray tube load. For a given thickness of anatomy, an optimum combination of spectrum kVp, contrast agent and supplemental filter was defined by maximum imaging performance for a minimum or tolerable x-ray tube load and patient dose. It was possible to both improve imaging performance and reduce dose by the use of an appropriate combination of spectrum kVp and filter. For gadolinium-based contrast, performance was optimized with tungsten filtration at 90 kVp for both thin and thick anatomies. It was not possible, however, to optimize the iodinated contrast performance with a single combination of supplemental filter and spectrum kVp. The optimal performance for iodinated contrast was achieved with gadolinium filtration at 60 kVp for thin anatomy and with ytterbium filtration at 80 kVp for thick anatomy. The best performance for thin anatomy was that of the combination of iodinated contrast/gadolinium filter at 60 kVp and the best performance for thick anatomy was that of the combination of gadolinium-based contrast/tungsten filter at 90 kVp.

There is a growing interest in performing intravascular interventions guided by MR imaging - a technique which offers the possibility of flow measurements during the intervention. For a reliable assessment of the haemodynamic significance of a stenosis, the flow and the pressure decay within the stenosis should both be measured. We have developed an optical, MR-compatible, pressure sensor (Annupres) that uses a novel annular element. Existing optical pressure sensors measure pressures unilaterally, thus giving rise to artefacts because of the dependence of the measurement on the angular orientation of the aperture. The annular element, however, measures blood pressure on all sides, and we show that by using circularly polarized light this pressure measurement is intrinsically insensitive to rotation of the sensor around its long axis. The Annupres sensor has been tested in an experimental set-up, and was able to measure pressures from 50 mmHg to 180 mmHg reliably with an accuracy of 1.5%.

We propose the use of anisotropic diffusion filtering to remedy difficulties in analysis of electronic portal images, stemming from their low contrast and high noise levels. Anisotropic diffusion is a nonlinear filter based on the numerical solution to the partial differential equation describing the process of diffusion. In this study we show that this filter is capable of greatly reducing noise in homogeneous areas of portal images while preserving the edges and contrast associated with anatomical features. We also demonstrate that the application of anisotropic diffusion leads to more consistent and reproducible visual extraction of features from portal images.

Stereotactic localization of an intracranial lesion by computed tomography or magnetic resonance imaging requires the use of a head frame that is fixed to the skull of the patient. To such head frames are attached either N-shaped or V-shaped localization rods. Because of patient positioning, the transverse imaging slices may not be parallel to the frame base; a coordinate transformation algorithm that takes this possibility into consideration is crucial. Here we propose such an algorithm for a head frame with V-shaped localization rods. Our algorithm determines the transformation matrix between the image coordinate system of a transverse image and the frame coordinate system. The determining procedure has three steps: (a) calculation of the oblique angles of a transverse image relative to the head frame and calculation of the image magnification factor; (b) determination of the coordinates of four central markers in both coordinate systems; and (c) determination of the 3×3 transformation matrix by using the coordinates of the four markers. This algorithm is robust in principle and is useful for improving the accuracy of localization.

The concept of field equivalence for electron beams is examined using a pencil beam theory applied to circular fields. It is shown that a circular field can be found for a field of any size, shape and energy for which the depth dose distribution is approximately equivalent. The usefulness of the concept in clinical dosimetry is discussed.

The derivation of the trigonometric equations necessary to calculate gantry, floor and collimator settings for a treatment plane at an angle ϕ to the transverse plane of the patient has been described previously. The derivation of a second set of equations to facilitate treatment in a plane at an angle ϕ to the coronal plane has also been described previously. This work reinterprets the geometry of inclined volumes and shows that essentially only one set of equations is required to determine the settings for treatment planes at an angle ϕ to either the transverse or coronal planes of the patient.

At an isocentric irradiation facility, the rotation axis of the treatment table has to be accurately aligned in vertical orientation to the isocentre, which is usually marked by three perpendicular laser planes. In particular, high precision radiotherapy techniques, such as radiosurgery or intensity modulated radiotherapy, require a higher alignment accuracy of the table axis than routinely specified by the manufacturers. A simple and efficient method is presented to measure the direction and the size of the displacement of the table axis from the isocentre as marked by the lasers. In addition, the inclination of the table axis against the vertical direction can be determined. The measured displacement and inclination provide the required data to correct for possible misalignments of the treatment table axis and to maintain its alignment. Measurements were performed over a period of two years for a treatment table located at the German heavy ion therapy facility. The mean radial distance between the table axis and the isocentre was found to be 0.25±0.25 mm. The mean inclination of the table axis in the XZ- and YZ-planes was measured to be -0.03±0.02° and -0.04±0.01°, respectively. The measurements demonstrate the good alignment of the treatment table over the analysed time period. The described method can be applied to any isocentric irradiation facility, especially including isocentric linear accelerators used for radiosurgery or other high precision irradiation techniques.

Radiochromic film is investigated for use in dosimetry in water phantoms as opposed to solid phantoms. Investigations are performed to measure the penetration rates of water into radiochromic film and to assess the effects on optical density that this penetration causes. The effects of film orientation during irradiation in water are also tested. Results show that only a small penetration rate is seen from water into the film which only affects the outer areas of the film, with penetration being less than 0.5 mm per hour. The optical density measurements of the film at 660 nm remain unchanged in the unaffected regions of the radiochromic film. Minimal effects are seen due to beam orientation in a water phantom as opposed to solid water phantoms in which an overestimation in dose is normally seen for parallel irradiation. Radiochromic film seems to be an adequate detector for dosimetry in a water phantom where high spatial resolution is needed and angle of beam incidence at the point of interest is important.

In the above paper, the authors omitted to make reference to an MSc Thesis on which some of the elements of the phantom design presented in the paper were based. In section 2.1 (Phantom design) the following statement should have been included.

The phantom developed was based on a design originally presented by Gilligan (1992).