Table of contents

Radiological images are increasingly being used in healthcare and medical research. There is, consequently, widespread interest in accurately relating information in the different images for diagnosis, treatment and basic science. This article reviews registration techniques used to solve this problem, and describes the wide variety of applications to which these techniques are applied. Applications of image registration include combining images of the same subject from different modalities, aligning temporal sequences of images to compensate for motion of the subject between scans, image guidance during interventions and aligning images from multiple subjects in cohort studies. Current registration algorithms can, in many cases, automatically register images that are related by a rigid body transformation (i.e. where tissue deformation can be ignored). There has also been substantial progress in non-rigid registration algorithms that can compensate for tissue deformation, or align images from different subjects. Nevertheless many registration problems remain unsolved, and this is likely to continue to be an active field of research in the future.

A rotation-based Monte Carlo (MC) simulation method (RMC) has been developed, designed for rapid calculation of downscatter through non-uniform media in SPECT. A possible application is downscatter correction in dual isotope SPECT. With RMC, only a fraction of all projections of a SPECT study have to be MC simulated in a standard manner. The other projections can be estimated rapidly using the results of these standard MC calculations. For efficiency, approximations have to be made in RMC with regard to the final scatter angle of the detected photons. Further speed-up is obtained by combining RMC with convolution-based forced detection (CFD) instead of forced detection (FD), which is a more common variance reduction technique for MC.

The RMC method was compared with standard MC for ^99mTc downscatter in a ²⁰¹Tl window (72 keV±10%) using a digital thorax phantom. The resulting scatter projections are in good agreement (maximum bias a few per cent of the largest value in the projection), but RMC with CFD is about three orders in magnitude faster than standard MC with FD and up to 25 times faster than standard MC with CFD. Using RMC combined with CFD, the generation of 64 almost noise-free downscatter projections (64×64) takes only a couple of minutes on a 500 MHz Pentium processor. Therefore, rotation-based Monte Carlo could serve as a practical tool for downscatter correction schemes in dual isotope SPECT.

Shuttling multileaf collimators (SMLCs) can increase the MU efficiency of intensity-modulated radiation therapy compared with the multiple-static-field (MSF-MLC) technique or dynamic MLC (DMLC) technique with conventional MLCs. In a previous paper (Phys. Med. Biol.45 3343-58) a particular SMLC was shown, for highly modulated intensity distributions, to increase the MU efficiency compared with the MSF-MLC technique. In this companion paper, two new arrangements similar to that described in the earlier paper, but with less mechanical complexity, are shown to be constructionally simpler but less MU efficient. Additionally another new concept of SMLC is shown which also increases the MU efficiency compared with the MSF-MLC technique and often improves the MU efficiency compared with the previously reported SMLC for highly modulated intensity distributions. It also leads to zero tongue-and-groove underdose in the direction orthogonal to that of the shuttling elements (so-called across-the-rows).

Monte Carlo simulations have been performed to determine the influence of collimator-scattered protons from a 150 MeV proton beam on the dose distribution behind a collimator. Slit-shaped collimators with apertures between 2 and 20 mm have been simulated. The Monte Carlo code GEANT 3.21 has been validated against one-dimensional dose measurements with a scintillating screen, observed by a CCD camera.

In order to account for the effects of the spatial response of the CCD/scintillator system, the line-spread function was determined by comparison with measurements made with a diamond detector. The line-spread function of the CCD/scintillator system is described by a Gaussian distribution with a standard deviation of 0.22 mm.

The Monte Carlo simulations show that protons that hit the collimator on the entrance face and leave it through the wall of the aperture make the largest scatter contribution. Scatter on air is the major contribution to the extent of the penumbra.

From the energy spectra it is derived that protons with a relative biological effectiveness greater than 1 cause at most 1% more damage in tissue than what would be expected from the physical dose.

A new real-time dose calculation and visualization was developed as part of the new 3D treatment planning tool OCTOPUS for proton therapy of ocular tumours within a national research project together with the Hahn-Meitner Institut Berlin. The implementation resolves the common separation between parameter definition, dose calculation and evaluation and allows a direct examination of the expected dose distribution while adjusting the treatment parameters. The new tool allows the therapist to move the desired dose distribution under visual control in 3D to the appropriate place. The visualization of the resulting dose distribution as a 3D surface model, on any 2D slice or on the surface of specified ocular structures is done automatically when adapting parameters during the planning process. In addition, approximate dose volume histograms may be calculated with little extra time. The dose distribution is calculated and visualized in 200 ms with an accuracy of 6% for the 3D isodose surfaces and 8% for other objects. This paper discusses the advantages and limitations of this new approach.

Boron neutron capture therapy (BNCT) is a radiation therapy in which the neutron capture reaction of ¹⁰B is used for the selective destruction of tumours. At the High Flux Reactor (HFR) in Petten, a therapy facility with an epithermal neutron beam has been built. In the first instance, patients with brain tumours will be treated. The doses delivered to the tumour and to the healthy tissue depend on the thermal neutron fluence and on the boron concentrations in these regions. An accurate determination of the patient dose during therapy requires knowledge of these time-dependent concentrations. For this reason, a γ-ray telescope system, together with a reconstruction formalism, have been developed. By using a γ-ray detector in a telescope configuration, boron neutron capture γ-rays of 478 keV emitted by a small specific region can be detected. The reconstruction formalism can calculate absolute boron concentrations using the measured boron γ-ray detection rates. Besides the boron γ-rays, a large component of 2.2 MeV γ-rays emitted at thermal neutron capture in hydrogen is measured. Since the hydrogen distribution is almost homogeneous within the head, this component can serve as a measure of the total number of thermal neutrons in the observed volume. By using the hydrogen γ-ray detection rate for normalization of the boron concentration, the reconstruction tool eliminates the greater part of the influence of the inhomogeneity of the thermal neutron distribution. MCNP calculations are used as a tool for the optimization of the detector configuration.

Experiments on a head phantom with 5 ppm ¹⁰B in healthy tissue showed that boron detection with a standard deviation of 3% requires a minimum measuring time of 2 min live time. From two position-dependent measurements, boron concentrations in two compartments (healthy tissue and tumour) can be determined. The reconstruction of the boron concentration in healthy tissue can be done with a standard deviation of 6%. The γ-ray telescope can also be used for in vivo dosimetry.

The relative output factors of the four helmets for a model B Leksell Gamma Knife and the end effect times for each helmet have been measured. For the three helmets with the smallest-diameter collimators a technique employing Kodak XV-2 film was used. The measured output factors are in good agreement with the values recommended by the manufacturer. The end effect times vary with the collimator size, with the shorter time occurring with the smaller collimator.

Boron neutron capture therapy (BNCT) is a form of targeted radiotherapy that relies on the uptake of the capture element boron by the volume to be treated. The treatment procedure requires the measurement of boron in the patient's blood. The investigation of a simple and inexpensive method for determining the concentration of the capture element ¹⁰B in blood is described here. This method, neutron flux depression measurement, involves the determination of the flux depression of thermal neutrons as they pass through a boron-containing sample.

It is shown via Monte Carlo calculations and experimental verification that, for a maximum count rate of 1×10⁴ counts/s measured by the detector, a 10 ppm ¹⁰B sample of volume 20 ml can be measured with a statistical precision of 10% in 32±2 min. For a source activity of less than 1.11×10¹¹ Bq and a maximum count rate of less than 1×10⁴ counts/s, a 10 ppm ¹⁰B sample of volume 20 ml can be measured with a statistical precision of 10% in 58±3 min. It has also been shown that this technique can be applied to the measurement of the concentration of any element with a high thermal neutron cross section such as ¹⁵⁷Gd.

For angiography using synchrotron radiation we measured the absorbed dose distribution in inhomogeneous phantoms with thin LiF:Mg, Cu, P, LiF:Mg, Ti thermoluminescent dosimeters (TLDs) in tissue and lung substitutes, and with Mg₂SiO₄:Tb TLDs in bone substitute for 33.32 keV monoenergetic photons from synchrotron radiation. The energy responses of the TLDs were measured in air for 10-40 keV monoenergetic photons. The values at 30 keV became smaller by 30% for LiF:Mg, Cu, P and larger by 22% for Mg₂SiO₄:Tb than the ratio of the mass energy absorption coefficients of the TLDs to that of air. These values were used to modify the calculated response of the TLDs in each phantom material. The absorbed dose distribution obtained was compared with that calculated using the Monte Carlo transport code EGS4 expanded to a low-energy region, and their agreement was confirmed taking linear polarization into account. In the bone substitute the dose increased by a factor of 3.9, while behind the bone the dose decreased drastically because of photon attenuation. In the lung substitute a slight dose difference from that in soft tissue was observed because of its different density. The LiF:Mg, Cu, P TLDs exhibited a better energy response, higher sensitivity and wider linear regions than did the other tissue-equivalent TLDs in the low-energy region.

The long-term sensitivity and calibration stability of liquid ionization chambers (LICs) has been studied at a local and a secondary standards dosimetry laboratory over a period of 3 years. The chambers were transported several times by mail between the two laboratories for measurements. The LICs used in this work are designed for absorbed dose measurements in the dose rate region of 0.1-100 mGy min^-1 and have a liquid layer thickness of 1 mm and a sensitive volume of 16.2 mm³. The liquids used as sensitive media in the chambers are mixtures of isooctane (C₈H₁₈) and tetramethylsilane (Si(CH₃)₄) in different proportions (about 2 to 1). Operating at a polarizing voltage of 300 V the leakage current of the chambers was stable and never exceeded 3% of the observable current at a dose rate of about 1 mGy min^-1. The volume sensitivity of the chambers was measured to be of the order of 10^-9 C Gy^-1 mm^-3. No systematic changes in the absorbed dose to water calibration was observed for any of the chambers during the test period (σ<0.2%). Variations in chamber dose response with small changes in the polarizing voltage as well as sensitivity changes with accumulated absorbed dose were also investigated. Measurements showed that the LIC response varies by 0.15% per 1% change in applied voltage around 300 V. No significant change could be observed in the LIC sensitivity after a single absorbed dose of 15 kGy. The results indicate that the LIC can be made to serve as a calibration transfer instrument and a reference detector for absorbed dose to water determinations providing good precision and long-term reproducibility.

Over several years the National Physical Laboratory (NPL) has been developing an absorbed dose calibration service for electron beam radiotherapy. To test this service, a number of trial calibrations of therapy level electron beam ionization chambers have been carried out during the last 3 years. These trials involved 17 UK radiotherapy centres supplying a total of 46 chambers of the NACP, Markus, Roos and Farmer types. Calibration factors were derived from the primary standard calorimeter at seven energies in the range 4 to 19 MeV with an estimated uncertainty of ±1.5% at the 95% confidence level. Investigations were also carried out into chamber perturbation, polarity effects, ion recombination and repeatability of the calibration process. The instruments were returned to the radiotherapy centres for measurements to be carried out comparing the NPL direct calibration with the 1996 IPEMB air kerma based Code of Practice.

It was found that, in general, all chambers of a particular type showed the same energy response. However, it was found that polarity and recombination corrections were quite variable for Markus chambers - differences in the polarity correction of up to 1% were seen. Perturbation corrections were obtained and were found to agree well with the standard data used in the IPEMB Code. The results of the comparison between the NPL calibration and IPEMB Code show agreement between the two methods at the ±1% level for the NACP and Farmer chambers, but there is a significant difference for the Markus chambers of around 2%. This difference between chamber types is most likely to be due to the design of the Markus chamber.

Lead shielding in contact with the patient's skin is often encountered in radiotherapy with electron beams. The influence of the lead shielding on dose distributions in the patient cannot fully be assessed using modern treatment planning systems. In this work the problem of quantifying the effect of lead shielding on dose distributions is addressed. Monte Carlo dose calculations were performed in a half-blocked water phantom shielded by lead, using a realistic model for the fluence of an electron linear accelerator. Electron beam energies of 6-20 MeV and lead thicknesses of 1-7 mm are used for 10×10 cm² and 5×5 cm² fields. The perturbation of the particle fluence and dose distributions in water introduced by the lead shielding is quantified. The effect of oblique electron beams on the dose perturbation is shown. A fictitious clinical example, the shielding of an eye in electron beam treatment, is used to demonstrate the usefulness of Monte Carlo based treatment planning algorithms that can incorporate the effects of lead shielding.

The objective of this study is to establish a comprehensive set of backscatter factors for mammography based on the exposure model proposed by the European Protocol on Dosimetry in Mammography. The Monte Carlo calculated backscatter factors (BSFs) presented in this study are for various exposure conditions encountered in mammographic practice as well as in calibration procedures. The data demonstrate the variation of the BSF as a function of the exposure parameters, hence enabling a better match with calibration conditions and, at the same time, reviewing the BSF data already recommended by the European Protocol. Furthermore, earlier data for BSF for general diagnostic radiology are validated.

In this work, polymer gel-MRI dosimetry (using VIPAR gels), radiographic film and a PinPoint ion chamber were used for profile measurements of 6 MV x-ray stereotactic beams of 5 and 10 mm diameter. The VIPAR gel-MRI method exhibited a linear dose response up to 32 Gy. VIPAR gels were found to resolve the penumbra region quite accurately, provided that the in-plane image resolution of the related T₂-map is adequate (⩽0.53 mm). T₂-map slice thickness had no significant effect on beam profile data. VIPAR measurements performed with a spatial resolution of 0.13 mm provided penumbra widths (80%-20% distance) of 1.34 and 1.70 mm for the 5 and 10 mm cones respectively. These widths were found to be significantly smaller than those obtained with the film (2.23 mm for the 5 mm cone, 2.45 mm for the 10 mm cone) and PinPoint (2.25 mm for the 5 mm cone, 2.52 mm for the 10 mm cone) methods. Regarding relative depth dose measurements, good correlation between VIPAR gel and PinPoint data was observed. In conclusion, polymer gel-MRI dosimetry can provide relatively accurate profile data for very small beams used in stereotactic radiosurgery since it can overcome, to some extent, the problems related to the finite size of conventional detectors.

Magnetization transfer and NMR relaxation rates were measured for water protons in two types of polymer gels developed for radiation dosimetry with MRI in order to quantify the contributions of different relaxation processes to the radiation response in such gels. Measurements included the rate of magnetization transfer between proton pools and the ratio of the sizes of exchanging pools, R₁ and R₂. A model of relaxation in irradiated gels is presented to explain their properties. The model incorporates three proton pools: free water, macromolecular and interfacial. Two pools are insufficient to model the data. In these systems, radiation-induced polymerization appears to increase the size of a solid-like macromolecular proton pool but does not affect the rate constant of magnetization transfer per proton from macromolecular protons to the free water protons. The relation between R₁ and the pool size ratio is consistent with free water exchanging with a macromolecular pool with an R₁ of ~8 Hz. In addition, the rate of magnetization transfer is not limited by the rate of chemical exchange between the free water and the interfacial protons, and magnetization transfer most probably occurs via labile proton exchange rather than via bound water molecules.

Dose profiles are presented resulting from computed tomography (CT). The profiles are positioned at the central axis, 1 cm away from the outer surface of the phantom, for single and multiple scans. A Hitachi W-1000 scanner is used with a thermoluminescent dosimeter (TLD), and standard dosimetry head and trunk phantoms. Regression equations are found linking the dose resulting from scattered radiation associated with a single scan to the distance from the scanning centre. The impact on the CT dose index value (CTDI) for varying integrating lengths is analysed. Some problems associated with CT dose measurement are noted, which may assist in the practical application of IBSS (International Basic Standard of Radiation Protection and Safety of Radiation Sources) guide levels.

The role of scatter in a cone-beam computed tomography system using the therapeutic beam of a medical linear accelerator and a commercial electronic portal imaging device (EPID) is investigated. A scatter correction method is presented which is based on a superposition of Monte Carlo generated scatter kernels. The kernels are adapted to both the spectral response of the EPID and the dimensions of the phantom being scanned. The method is part of a calibration procedure which converts the measured transmission data acquired for each projection angle into water-equivalent thicknesses. Tomographic reconstruction of the projections then yields an estimate of the electron density distribution of the phantom. It is found that scatter produces cupping artefacts in the reconstructed tomograms. Furthermore, reconstructed electron densities deviate greatly (by about 30%) from their expected values. The scatter correction method removes the cupping artefacts and decreases the deviations from 30% down to about 8%.

Hot electrons confined in a compact magnetized plasma (in a 1 litre chamber) and heated by electron cyclotron resonance were used to provide an intense reproducible x-ray emission. Since electrons were generated in situ, the use of filaments and high voltages at the base of the x-ray tubes was avoided. The source can be pulsed or operated in a highly stable continuous mode. The spatial and energetic characteristics of the hot electrons were determined via their bremsstrahlung by comparison with theoretical calculations. In this test x-ray source the average energy of the hot electron flux was in the range 10-40 keV, depending on the operating parameters: intermagnet distance (6-9 cm), power of the 2.45 GHz microwave (200-1000 W), chamber pressure (4×10^-5 - 9×10^-5 Torr). The insertion of solid targets into the electron flux leads to various focal spot sizes, enabling application in radiology and mammography. The images of a mammographic phantom were obtained in a few seconds with good contrast (microcalcifications being successfully detected).

An iterative Bayesian reconstruction algorithm for limited view angle tomography, or ectomography, based on the three-dimensional total variation (TV) norm has been developed. The TV norm has been described in the literature as a method for reducing noise in two-dimensional images while preserving edges, without introducing ringing or edge artefacts. It has also been proposed as a 2D regularization function in Bayesian reconstruction, implemented in an expectation maximization algorithm (TV-EM). The TV-EM was developed for 2D single photon emission computed tomography imaging, and the algorithm is capable of smoothing noise while maintaining edges without introducing artefacts.

The TV norm was extended from 2D to 3D and incorporated into an ordered subsets expectation maximization algorithm for limited view angle geometry. The algorithm, called TV3D-EM, was evaluated using a modelled point spread function and digital phantoms. Reconstructed images were compared with those reconstructed with the 2D filtered backprojection algorithm currently used in ectomography. Results show a substantial reduction in artefacts related to the limited view angle geometry, and noise levels were also improved. Perhaps most important, depth resolution was improved by at least 45%. In conclusion, the proposed algorithm has been shown to improve the perceived image quality.

The purpose of this paper is to compare two models of the electrical properties of tissue, which may be used to relate the effective conductivity to the volume fraction f of cells in the tissue. Both models assume that tissue comprises spherical cells, which behave electrically as dipoles. The first model, developed by Hanai, describes the tissue as a concentrated suspension of weakly conducting spheres in a conducting medium, with each sphere experiencing a uniform mean field. The second approach, developed by Chiew and Glandt, explicitly describes the effect of a random but statistically homogeneous cell structure on the average field and magnitude of the dipole interaction. The two analyses are identical to first order in f, but differ in the way in which the interactions between the dipoles are accounted for. The model developed by Chiew and Glandt appears to offer a more robust theoretical framework for describing the electrical properties of tissue. The comparison aims to contribute to an improved understanding of the relationship between the electrical properties and spatial structure of tissue.

A theoretical approach is presented to determine absorption changes in different compartments of a layered structure from distributions of times of flight of photons. In addition resulting changes in spatial profiles of time-integrated intensity and mean time of flight are calculated. The capability of a single-distance, time-domain method to determine absorption changes with depth resolution is tested on a layered phantom. We apply this method to in vivo measurements on the human head (motor stimulation, Valsalva manoeuvre) and introduce a small-sized time-domain experimental set-up suitable for bedside monitoring.

Our objective was to appraise whether an increased amniotic fluid pressure by polyhydramnios can beneficially affect monochorionic twins that are haemodynamically connected by arterio-venous plus arterio-arterial placental anastomoses. We assessed the effects of polyhydramnios in monochorionic twin placentas, combining (a) data from previous in vitro placental perfusion experiments in singleton term placentas under simulated normal and increased amniotic fluid pressures with (b) logical deduction from observations made in monochorionic twins. Our hypothesis is that in monochorionic placentas, an increased amniotic fluid pressure increases the placental microvascular resistance but not the resistance of placental chorionic plate arteries. Hence, an increased amniotic fluid pressure increases the microvascular resistance of the joint cotyledon, the arterio-venous resistance, but not the arterio-arterial resistance. This proposed mechanism reduces arterio-venous but not oppositely directed arterio-arterial transfusion. Therefore, reversal of the normal direction of net foeto-foetal transfusion may develop, which will reduce the circulatory imbalance that evolved between the monochorionic foetal twins. In contrast, in monochorionic twins connected by unidirectional or bidirectional arterio-venous anastomoses reversal of the normal direction of net foeto-foetal transfusion will not occur. In conclusion, reversal of the normal direction of net foeto-foetal transfusion, induced by polyhydramnios, is protective against the onset and severity of twin-twin transfusion syndrome in monochorionic twins connected by arterio-venous plus arterio-arterial anastomoses, but not by unidirectional or bidirectional arterio-venous anastomoses.

Though the optics of red blood cells as well as whole blood has been studied extensively, an effective scattering phase function for whole blood is still needed. The interference of waves scattered by neighbouring cells cannot be neglected in highly concentrated suspensions such as whole blood. As a result, the phase function valid for single erythrocytes may fail to describe a single scattering process in whole blood with physiological haematocrit (Hct ≈ 0.4). In this study we compared the results obtained in goniophotometric measurements of blood samples with the results of angle-resolved Monte Carlo simulations. The results show that a Henyey-Greenstein phase function with an anisotropy factor of 0.972 is an adequate approximation for the effective scattering phase function of whole blood with high haematocrit at a wavelength of 514 nm.

Samples with 5.1%, 9.8% and 15.3% HbA1c were extracted from normal subjects and patients with slight and serious diabetes respectively. Extended x-ray absorption fine structure spectra of Fe K absorption were collected at the EXAFS experimental station of the Beijing Synchrotron Radiation Facility. The step-by-step fluorescent mode was employed with a count time of 10 s per point. Several independent scans were averaged to eliminate the statistical noise. Reference backscattering amplitudes and phaseshifts were calculated using the curve wave theory (FEFF code) of EXAFS. Apart from the nitrogen neighbours around the central iron atom, oxygen neighbours are also found. The Fe-N bond length increases by about 0.02 Å for the sample with 15.3% HbA1c compared with the others, but the Fe-O bond length is almost unchanged. With increasing of HbA1c concentration, the content of Hb increases and the content of HbO₂ decreases. This demonstrates that the glycosylation of haemoglobin will decrease its ability to carry oxygen.

In this study a complete set of dosimetric data is presented for the high dose rate (HDR) source from Amersham used in the Buchler remote afterloading HDR unit. These data have been calculated by means of the Monte Carlo simulation code GEANT taking into account the detailed geometry of the source. Absolute dose rate distributions in water were calculated around this source and are presented as conventional 2D Cartesian look-up tables. All dosimetric quantities recommended by the AAPM Task Group 43 report have been calculated. Quantities determined are: dose rate constant, radial dose function, anisotropy function, anisotropy factor and anisotropy constant. The dose rate distributions of the Buchler HDR source are compared with those of other HDR sources used in brachytherapy, showing that the differences are large in zones near the long source axis due to oblique filtration. These Monte Carlo simulated data in water can be used for clinical applications.

In the above rapid communication, the name of one of the authors has been omitted. The author in question is K W Burn, who should have been the final author in the article, and who gave a strong contribution to the development of the Monte Carlo code and performed the Monte Carlo simulation of the Mevatron Siemens accelerator. The full list of authors and affiliations should read as follows.

Carla Ongaro¹, Alba Zanini², U Nastasi³, José Ródenas⁴, Giuseppe Ottaviano¹, Claudio Manfredotti¹ and Ken William Burn⁵

¹Dipartimento di Fisica Sperimentale, Università di Torino, Via P Giuria 1, 10125 Torino, Italy ²INFN-sezione di Torino, Via P Giuria 1, 10125 Torino, Italy ³Ospedale S Giovanni AS, Via Cavour 30, 10133 Torino, Italy ⁴Departamento de Ingenieria Quimica y Nuclear, Universidad Politécnica de Valencia, Apartado 22012, E-46071 Valencia, Spain ⁵ENEA ERG, Via Martiri di Monte Sole 4, 40129 Bologna, Italy

ed J R Williams and D I Thwaites Oxford: Oxford University Press (2000) 332pp, price: £34.50, ISBN: 0-19-262878-X

The second edition of Radiotherapy Physics: in Practice is, like its predecessor, intended as a practical guide particularly for the trainee medical physicist. It has been very successful at fulfilling its intention of being a comprehensive review of the practice of radiotherapy physics by providing a good introduction to its various topics, without requiring any extensive physics or mathematics. It is also a useful text for more experienced medical physicists, dosimetrists and medical technical officers. Some of the sections would also interest radiographers and clinical oncologists.

Since the first edition was published in 1993 there have been significant advances in radiotherapy technology and these have been adequately covered in the second edition. A number of chapters concerning the established basis of radiotherapy physics have remained basically unchanged from the first edition.

The chapter on 'Simulation and imaging for radiation therapy planning' now includes target definitions relating to ICRU Report 50 and has an excellent section on the CT simulator and virtual simulation. The original chapter on external beam treatment planning has been expanded into three chapters with a separate section dedicated to electron beam planning. An additional chapter covering the changes in treatment planning initiated by the advances in technology - multileaf collimators, electronic portal imagers and dynamic therapy - gives a good practical background to these topics without the in-depth mathematics found in scientific papers. The improvements in CT-based and 3D treatment planning such as automatic outline and target drawing, beam's eye view, digitally reconstructed radiographs and dose-volume histograms are also covered in some detail and are a welcome addition in this second edition.

Some of the significant changes in dosimetry in the last few years have also been included in their relevant sections.

The final chapter on 'Quality assurance in radiotherapy physics' is an excellent review on the important topic of quality assurance management or quality systems, and how they are related to all the activities in a radiotherapy physics department which affect the accuracy of radiation treatment in the clinical oncology department. The structure of the quality system, how it is implemented in practice and the process of certification are all explained. The chapter concludes with a discussion on dosimetry audits.

The book is produced to a high standard throughout with 139 figures, including many very clear line drawings. This book should be in the library of radiotherapy physics departments and, at only £34.50 for 332 pages, is certainly affordable. I would recommend it to any medical physics trainee as a text to provide them with guidance for the many practical problems in radiotherapy physics, and as a source of useful references giving further background theory on these topics.