Table of contents

Unidirectional blood-to-brain and blood-to-tumour transport rate constants (K₁) for ⁸²Rb (half-life 76 s) and plasma water volume per unit mass of brain/tumour tissue (V_p) can be estimated in vivo using dynamic positron emission tomography (PET). The accuracy of these estimates depends upon the accuracy of PET measurements of regional brain/tumour radioactivity and scintillation well detector measurements of whole-blood radioactivity, which, in turn, depend upon the time course of arterial blood radioactivity. A two-compartmental model has been employed to derive estimates for K₁, k₂ (efflux rate constant) and V_p from ⁸²Rb/PET data. Errors in these parameter estimates have been studied (1) qualitatively using sensitivity function analysis and (2) quantitatively using computer simulations. The effect of adding a third irreversible compartment and its unidirectional rate constant, k₃, has also been investigated. The advantages and disadvantages of bolus injection vs continuous infusion protocols are discussed.

For pt.I see ibid., vol.34, no.12, p.1773-84 (1989). Computer simulations and error analysis of a simple two-compartment, passive-diffusion exchange model for ⁸²Rb across the blood-brain-barrier (BBB) have demonstrated the feasibility of obtaining useful estimates of unidirectional rate constant (K₁) and tissue-blood water volume (V_b) in vivo using dynamic positron tomography (PET). The coefficients of variation (CV) in parameter estimates for 20 studies on ten patients are shown to have mean values of 10% for tumour K₁ and V_b, 6% for normal brain V_b and 30% for normal brain K₁. Ten test-retest studies show a very high correlation (R²>0.88) between the estimated parameter values. Apparent tissue blood volume (V_b) estimates obtained from ⁸²Rb studies underestimate the blood volumes obtained by single-breath C¹⁵O studies by approximately 15%. The data from these clinical studies are in agreement with the predictions of the computer simulations and suggest the suitability of ⁸²Rb/PET in the quantification of BBB functional changes.

A technique is suggested for quantitative estimation of contrast medium in human organs using digital subtraction fluorography. A reference wedge of tin is placed in the X-ray field for one of two mask images. The subtracted sequence then contains a set of reference pixel values which can be used as calibration data for the administered contrast medium. The factors affecting quantitative measurement are discussed and experimental estimates of errors made. The characteristics of veiling glare in the system were measured and the effectiveness of a deconvolution algorithm demonstrated. Pixel values corresponding to various thicknesses of a tin reference wedge were compared with those corresponding to thicknesses of tin in a test wedge positioned within a phantom to simulate the patient. Differences between pixel values were measured for variations in several radiographic parameters. Frame-to-frame changes in X-ray output were minimised by using the modal pixel value in the image as a background reference. An estimation of mass thickness of known iodine solutions was made using the technique. In the range 0.5-20.0 mg cm^-2 iodine, the mean estimated value was 0.8 mg cm^-2 lower than nominal, with an uncertainty of 1.1 mg cm^-2.

A Monte Carlo simulation has been used for the determination of the gap correction to be applied to the calorimetric measurement of absorbed dose in graphite in a ⁶⁰Co beam. The results show that this correction, which has been neglected so far by most national standards laboratories, is not negligible and depends on many parameters. A comparison with the available experimental data shows a general good agreement.

An alpha irradiation device is described that utilises a commercially available disc-shaped americium-241 source of 8 cm diameter. The alpha particles traverse a moving collimator and the source is rotated to reduce the influence of source inhomogeneities. Source, collimator and a shutter disc are mounted in a container which is flushed with helium to reduce energy losses of the alpha particles before reaching the exit foil. The shutter disc is activated by a computer-controlled step motor. The broad beam of alpha particles emerges from the exit window of the container with a remaining range in tissue of about 15 mu m. An intermittent computer-controlled use of a preabsorber makes it possible to reduce dose differences within a depth up to 12.5 mu m to not more than +or-3%. With the commercially available americium sources a dose rate of 0.2 Gy min^-1 is reached; this can be increased by utilising a somewhat wider collimator.

In vivo bone lead measurements have been made on a group of about 120 people, most of whom were lead exposed workers. Two different X-ray fluorescence (XRF) techniques were used to make measurements at three bone sites. Finger lead was measured using ⁵⁷Co sources, and lead measurements were made in both tibia and calcaneus with a technique based on ¹⁰⁹Cd sources. The results of the bone lead measurements correlated strongly with each other and with the index of cumulative exposure, thus confirming the value and reliability of these in vivo measurements as a tool in the study of chronic lead exposure. Measurement precision, +or-1 standard deviation, was highest for tibia +or-7.4 mu g (g bone mineral)^-1, +or-16.6 mu g (g bone mineral)^-1 for the calcaneus and lowest for phalangeal lead +or-25.0 mu g (g bone mineral)^-1. Maximum absorbed doses to the skin were comparable for all three measurements (1-3 mGy). The mean whole body dose equivalents were all low, but that for the finger measurement, 0.1 mu Sv, was significantly less than for the calcaneus and tibia measurements 3-5 mu Sv.

The influence of energy and angular spread and electron and photon contamination on the water/air stopping-power ratios for 'realistic' electron beams of 10 MeV has been investigated using the Monte Carlo method. Differences smaller than 0.5% have been found in the s_W,air value at the depth of maximum absorbed dose compared with s_W,air(E₀,z) determined according to most dosimetry protocols. The use of broad independent energy and angular distributions to sample the initial state of electrons for Monte Carlo simulations has been analysed. Uncertainties in s^SA_W,air values, evaluated with a Monte Carlo iterative technique, are approximately 0.5% at d_max. The combination of uncertainties in s^SA_W,air due to the calculation procedure and to the 's_W,air(E₀,z) method' yields an estimated total uncertainty of approximately 1% for the s_W,air at d_max in clinical electron beams with energies around 10 MeV, which is smaller than values quoted recently.

Describes how an experimental arrangement for the calibration of Fricke solution in terms of absorbed dose to water can be utilised to determine total, i.e. collisional and radiative, mass stopping power of high-energy electrons. As a first result the measurement of the total mass stopping power of polystyrene at about 5.3 MeV kinetic electron energy is presented in detail. Comparison of the obtained value with the corresponding result of recent theoretical computations shows agreement within the measurement uncertainty of about 1.2% (SD).

The backscatter factors computed by Grosswendt (1984) for X-rays generated at voltages between 10 and 100 kV using a Monte Carlo method differ significantly from the corresponding factors published in supplement 17 of the British Journal of Radiology (1983). The data may be subject to systematic uncertainties and it was therefore thought necessary to undertake a study of the backscatter in a uniform experiment in order to verify these data and the theoretical calculations of Grosswendt (1984). A special 3 mm thin plane-parallel ionisation chamber was constructed for this work and measurements were performed using water as a scattering medium. The experiments involved field sizes of 2.0 to 20 cm beam diameter and X-ray qualities between 0.1 and 4 mm Al for generating voltages between 16 and 140 kV. The results yielded backscatter factors which are in close agreement with those obtained by Monte Carlo calculations. The agreement with the supplement 17 data is only satisfactory at HVLs above 1 mm Al. The possible causes for the disagreement between the various published data are analysed.

The Erlangen micro-lightguide spectrophotometer EMPHO I was designed for fast diffuse reflection (remission) spectrophotometry in small tissue volumes. The aim was to construct a compact, modular instrument with a high repetition rate which can be adapted to moving organs, e.g. the beating heart in situ, by the use of highly flexible micro-lightguides. Focusing problems, which cannot be solved when conventional optical devices such as microscopes are used in moving tissues, become negligible. A bandpass interference filter disk, which is rotated by a motor serves as a monochromating unit. One diffuse reflection spectrum in a selected wavelength domain is recorded during each revolution of the motor. Special filter disks, with spectral ranges of 400-520, 500-630, 600-1200 nm can be used for different tasks. The monochromated light is transmitted by means of a flexible fluid-lightguide to a photomultiplier tube. The electrical signal, which is proportional to the light intensity is recorded by an IBM-compatible AT. An analogue to digital converter has been developed for the AD conversion.

Remission spectroscopy in living tissues can only be performed because scattering processes induce a pronounced amount of backscattered light. New types of scattering chambers were constructed in order to investigate the angular distribution of light intensities caused by scattering events typical for living tissues. Different solutions and suspensions containing microspheres, haemoglobin molecules, erythrocytes and liver homogenate were used in order to gain basic information applicable to remission spectrophotometry in tissue. Reflection spectroscopy in biological tissues is only possible because of the scattering properties of the material under observation. Light which penetrates the sample is remitted from the tissue. The remission is caused by the scattering of the light by different biological structures and particles. This scattered light gives information about the optical properties of the particles and structures in an integral form. A scattering chamber for small sample volumes has been developed for measurements of the optical properties of various particles in suspension.

The authors designed and tested two microwave applicators suitable for intracavitary hyperthermia treatment of advanced cancer colli uteri in combination with ionising radiation. The antennas operate at the frequencies of 2450 MHz and 680 MHz. Hence patients with differing extents of tumour growth along the collum uteri can be treated. The antenna is covered with a PTFE tube to form an applicator. The absorption in the tube is very low, and consequently the power is delivered outside the tube.

The availability in the 1980s of precision spectrophotometers has rekindled interest in Fricke dosimetry as a research tool. The ferric-ion yield, G(Fe⁺³), is known to depend on the temperature of the dosemeter during irradiation. The results reported here represent a new determination of the temperature dependence of G(Fe³⁺) using precision spectrophotometry and a state-of-the-art Fricke dosimetry system (Ross et al., 1989).

Intralipid is a fat emulsion which is widely used as the basis of phantom materials for the investigation of the optical properties of tissues, a subject of recent increased interest due to a growth in the use of photodynamic therapy as a treatment of cancer. In the course of the present study it was noted that the effective attenuation coefficient and therefore the light penetration depth of an Intralipid suspension was not directly proportional to the concentration. The added absorber method was used to determine the absorption and reduced scattering coefficients of Intralipid for any concentration in aqueous solution. The results show that most of the absorption is due to the suspending medium, water.

In test object measurements of slice profiles in MR imaging the two methods usually utilised (ramp and wedge techniques) each have their own advantages and disadvantages. Whereas the ramp method yields immediate slice profiles from the images it is only capable of accuracy (errors 5% or less) where the plate thickness is more than a factor of 2.5 to 3 times less than the FWHM slice thickness. This leads to problems for thin slices, particularly since the effects of noise become more severe for thin plates. In the case of the ramp method no difficulties exist with thin slices except that the requirement for numerical differentiation leads to an increased sensitivity to noise. For the same slice thickness the wedge method is approximately twice as sensitive to noise as the ramp method using a 1 mm plate. In both methods it is highly desirable to average several lines in the image to obtain a profile-a facility which usually does not exist in commercial imaging apparatus.

When measured absorbed doses for photon beams are to be separated into their primary and scatter components, the extrapolation method proposed gives consistent results. Information about the head-scatter factor H(s) is required to implement the extrapolation and can be determined by supplementary measurements in air. In this study, H(s) was derived from normalised peak scatter factor for the X-ray beams and already eliminated in the ⁶⁰Co tissue-air ratio data. Information about the attenuation of the beam is also needed and can be provided by narrow-beam measurements in air. The procedure can be applied to tissue-maximum ratio data, as was done in this study, but also to other representations of doses on the central axis, after transformation to TMR. All doses used for the extrapolation must be measured under electron equilibrium conditions. It simplifies the procedure if all tabulated doses are normalised to one combination of depth and field size, also sufficient for electron equilibrium.

Owing to the known potential of maximum likelihood reconstruction, the authors have developed a cone beam single photon emission computerised tomography maximum likelihood algorithm in an effort to reduce or remove the reconstruction distortions in cone beam filtered back projection. Previous work (Floyd et al., 1986) was done on cone beam maximum likelihood, but with poor impractical resolution because of practical limitations on computer memory. For the present study, the authors have developed a new implementation which provides higher, potentially practical, resolution by recalculating all matrix elements as needed rather than calculating once and storing in the computer.