Purpose Determining treatment margins for inter-fractional motion of moving and deformable clinical target volumes (CTVs) remains a major challenge. This paper describes and applies an optimisation algorithm designed to derive such margins. Material and methods The algorithm works by expanding the CTV, as determined from a pre-treatment or planning scan, to enclose the CTV positions observed during treatment. CTV positions during treatment may be obtained using, for example, repeat CT scanning and/or repeat electronic portal imaging (EPI). The algorithm can be applied to both individual patients and to a set of patients. The margins derived will minimise the excess volume outside the envelope that encloses all observed CTV positions (the CTV envelope). Initially, margins are set such that the envelope is more than adequately covered when the planning CTV is expanded. The algorithm uses an iterative method where the margins are sampled randomly and are then either increased or decreased randomly. The algorithm is tested on a set of 19 bladder cancer patients that underwent weekly repeat CT scanning and EPI throughout their treatment course. Results From repeated runs on individual patients, the algorithm produces margins within a range of ±2 mm that lie among the best results found with an exhaustive search approach, and that agree within 3 mm with margins determined by a manual approach on the same data. The algorithm could be used to determine margins to cover any specified geometrical uncertainty, and allows for the determination of reduced margins by relaxing the coverage criteria, for example disregarding extreme CTV positions, or an arbitrarily selected volume fraction of the CTV envelope, and/or patients with extreme geometrical uncertainties. Conclusion An optimisation approach to margin determination is found to give reproducible results within the accuracy required. The major advantage with this algorithm is that it is completely empirical, and it is therefore particularly useful for CTVs where the geometrical uncertainties are difficult to model, such as the bladder.

A method is described for calculating the output from conformally shaped megavoltage X-ray beams. The model has been developed for Varian accelerators but is shown to work for accelerators from another manufacturer. The use of dynamic wedging and both static and dynamic multileaf collimated beams are included in the model. For any linear accelerator, the data required are a set of measured output factors for square beams, an in-air profile and a limited number of readily available parameters defining the geometry of the head of the accelerator. The three components of the output, namely primary, head scatter and phantom scatter are modelled and calculated individually for any point in a beam. An optimization procedure is developed that automatically determines the eight parameters required to model an accelerator in order for these calculations to be performed. The performance of the method is demonstrated for shaped beams using asymmetric and multileaf collimation, both with and without wedging, and for a range of beam energies. The model has been incorporated into a computer program that is used clinically.

Background and purposeDifferent treatment margins and fractionation schedules are used in conformal radiotherapy (CRT) of urinary bladder cancer. This study compared intestine and rectum dose–volume histogram (DVH) data and normal tissue complication probability (NTCP) estimates for various clinically applied margins and fractionation schedules in bladder irradiation.Patients and methodsNormal tissue dose distributions in fifteen bladder cancer patients treated with CRT were studied using standard three- and four-field configurations. The impact of margin width on intestine and rectum dose distributions was initially evaluated using DVH data. NTCP modelling with the probit model was used to compare the impact of choice of margin size and fractionation schedule. The analysis included margin combinations of 1.0 cm isotropic (narrow margins) and 1.2–2.0 cm non-isotropic (wide margins) and fractionation schedule alternatives of 52.5Gy/20, 55Gy/20, 57.5Gy/20 and 64Gy/32.ResultsUsing wide as compared to narrow margins, the volumes of intestine and rectum receiving high doses increased by factors of approximately two and four, respectively. Similar differences between wide and narrow margins were found when calculating intestine and rectum NTCPs. The impact of margin size depended strongly on the volume effect expressed by the NTCP model parameters. With standard parameters, however, the choice of margins and fractionation schedule had a similar impact on intestine NTCPs, while for the rectum, the choice of margin had a greater impact than the choice of fractionation. For a given margin size, the intestine and rectum NTCPs for the 55Gy/20 and the 64Gy/32 schedules were comparable. For clinics using narrow margins and a fractionation of 52.5Gy/20, the NTCP modelling suggested that a change in fractionation schedule (to 55Gy/20 or 64Gy/32) or a change to wide margins would have a similar effect on the intestine NTCP predictions.ConclusionsThis modelling study documented that the choice of margins was as important as the choice of fractionation in terms of intestine and rectum DVH data and NTCP predictions.

Abstract Treatment planning is the process whereby the therapeutic strategy of the radiation oncologist is realized as a set of treatment instructions together with a physical description of the distribution of dose in the patient.

Background and purpose: Dose heterogeneity in tangential breast irradiation has been shown to be as high as 20% and may lead to problems in local control and cosmesis. In this study, dose heterogeneity in three dimensions (3D) in the breast irradiated with wedged tangential beams is assessed and the improvement which can be made by the use of individualised two dimensional (2D) compensators is established. The compensation required is calculated in two ways: (I) by an iterative technique giving a uniform dose on a plane through the isocentre normal to the central axis of each beam, and (II) by inverse planning using an optimisation technique based on simulated annealing.Materials and methods: A total of 17 patients with histologically proven T0-3, N0, N1, M0 breast cancer undergoing breast irradiation following wide local excision, were CT scanned using contiguous 1 cm slices from approximately 2 cm superior to 2 cm inferior of the irradiated volume. The dose distributions are determined using a 3D algorithm that calculates primary and scatter dose separately using a differential scatter air ratio method and corrects both for the presence of heterogeneities. The iterative technique achieves a dose variation of better than 0.5% on the plane through the isocentre with compensation on both beams. Compensation for the lateral beam only is calculated using the optimisation technique in order to minimise the scatter dose to the contralateral breast. The optimisation algorithm minimises the dose variance over the target and sets upper dose limits for the lung and the remainder of the irradiated volume.Results: For the group of patients the average dose heterogeneity in 3D using wedges is 12% (range 8-17%), which reduces to 8% (5-16%) using compensation on a plane and to 5% (4-7%) using the optimisation technique.Conclusions: Inverse planning is normally used for complex radiotherapy techniques but when applied to tangential breast irradiation, can reduce the dose heterogeneity through the breast as a whole to as little as 4%, with potential benefits in local control and cosmesis. (C) 1999 Elsevier Science Ireland Ltd. All rights reserved.

Background and purpose: Dose heterogeneity in tangential breast irradiation has been shown to be as high as 20% and may lead to problems in local control and cosmesis. In this study, dose heterogeneity in three dimensions (3D) in the breast irradiated with wedged tangential beams is assessed and the improvement which can be made by the use of individualised two dimensional (2D) compensators is established. The compensation required is calculated in two ways: (I) by an iterative technique giving a uniform dose on a plane through the isocentre normal to the central axis of each beam, and (II) by inverse planning using an optimisation technique based on simulated annealing. Materials and methods: A total of 17 patients with histologically proven T0-3, N0, N1, M0 breast cancer undergoing breast irradiation following wide local excision, were CT scanned using contiguous 1 cm slices from approximately 2 cm superior to 2 cm inferior of the irradiated volume. The dose distributions are determined using a 3D algorithm that calculates primary and scatter dose separately using a differential scatter air ratio method and corrects both for the presence of heterogeneities. The iterative technique achieves a dose variation of better than 0.5% on the plane through the isocentre with compensation on both beams. Compensation for the lateral beam only is calculated using the optimisation technique in order to minimise the scatter dose to the contralateral breast. The optimisation algorithm minimises the dose variance over the target and sets upper dose limits for the lung and the remainder of the irradiated volume. Results: For the group of patients the average dose heterogeneity in 3D using wedges is 12% (range 8±17%), which reduces to 8% (5±16%) using compensation on a plane and to 5% (4±7%) using the optimisation technique. Conclusions: Inverse planning is normally used for complex radiotherapy techniques but when applied to tangential breast irradiation, can reduce the dose heterogeneity through the breast as a whole to as little as 4%, with potential bene®ts in local control and cosmesis. q 1999 Elsevier Science Ireland Ltd. All rights reserved.

Techniques for reducing computation time in 3D photon dose calculations are addressed with specific emphasis given to the convolution/superposition approach. A single polyenergetic superposition model calculating absorbed dose per incident photon fluence (Gy cm2) was developed in terms of TERMA and a total energy deposition kernel (a total point spread function). A novel approach was devised for reducing calculation time. The method, named the CF method, was based on the use of a conventional, fast model (here a modified power-law method was used) for the generation of 3D dose distributions on a fine dose matrix. Superposition calculations were carried out on a coarse matrix and calculation speed was increased simply by reducing the number of calculations. A set of correction factors was derived on the coarse grid from the ratio of the dose values from superposition to those from the conventional algorithm. These were interpolated onto the fine matrix and used to modify the dose calculation from the conventional algorithm. The method was tested in a worst-case example where large dose gradients were present and in a clinically relevant irradiation geometry. It is shown that the time required for the generation of a 3D matrix with superposition can be reduced by at least a factor of 100 with no significant loss in accuracy.

A comprehensive beam model for three-dimensional radiotherapy planning is described. Input of patient data is either from an arbitrary spaced set of computed tomography (CT) scans throughout the treatment volume or from outlines traced on a digitizer. Any gantry, head or table rotation is allowed as well as any beam portal shape including internal blocking. Account is taken of beam modification due to the presence of the beam flattening filter, wedges and compensators. Both portal image and beam's eye view displays are readily available. A dose calculation algorithm is presented which uses a differential scatter-air ratio method to calculate the scatter dose in the presence of inhomogeneities. The beam model and algorithm form the basis of an interactive three-dimensional planning system running on a UNIX workstation, and performance times for dose calculation are given.

Physicists, radiotherapists and radiographers have worked together to enhance the quality and accuracy of radiotherapy for tangential irradiation of the chest wall and breast. Each stage of the process has been reassessed and improved. A technique has been developed on the treatment simulator which determines the required beam directions and sizes in a straightforward manner. A computerised tomography facility has also been developed on the simulator and this provides one or more slices for planning through the treatment volume, thus allowing accurate determination of external contour and lung position with the patient in the treatment position. The beam edge entry points and the isocentre position can be seen from radiation opaque markers placed on the skin surface, allowing accurate reproduction of the treatment beam positions as set on the simulator. A photon beam algorithm that corrects for the changes in scatter dose in a 3-dimensional (3-D) inhomogeneous situation has been developed and applied to tangential chest wall irradiation. This has shown large differences (up to 10%) in dose compared to a conventional 2-dimensional algorithm. The changes in dose distribution due to the accurate determination of lung position have also been investigated. A method has been developed of measuring the volume of ipsilateral lung irradiated. Dose-area histograms are used to evaluate the fraction of the lung area irradiated in the central slice, and an estimate of the volume irradiated can be made using a beam's eye radiograph and the simulator CT image.

A new 3-dimensional scatter algorithm based on the principles of the equivalent tissue air ratio (ETAR) method has been developed. The algorithm models the relative amount of scattered radiation reaching a point in both homogeneous and inhomogeneous situations, and derives a correction factor which can be applied to the scatter component of the dose to the point. The algorithm has been installed on an in-house developed planning system running on a Sun Sparcstation and its implementation and performance are described. The new algorithm has been evaluated by comparison with measurement in experimental situations designed to provide a rigorous test of the model. Comparisons with ETAR are also made, and the new algorithm gives a considerable improvement in accuracy over ETAR without any significant increase in calculation time.

The use of a radiotherapy simulator linked to a treatment planning computer to produce computed tomographic (CT) images for radiotherapy planning was described by Redpath and Wright (1985). The outlines of the external body contour and any inhomogeneous structures are extracted automatically from the CT image by a contouring algorithm available in treatment planning software. Experimental results with phantoms have shown this technique to be accurate to ±5 mm. Prior to the availability of CT images, the standard method of obtaining body outlines was to use the range finder on the simulator to measure the focus–skin distance at 10° intervals around the patient. These values were typed into a computer which produced a life-sized outline approximated by a polygon. The method was difficult to use when the range finder was pointing up from below the horizontal, and obviously gives no information on the positions of internal structures which have to be estimated from radiographs. Again, experiments with phantoms have shown this method to be accurate to ±20 mm in determining the body contour in the region accessible to the range finder. The contour was completed by relating its position with respect to the simulator couch top through knowledge of the depth of the isocentre and the antero-posterior separation of the patient.

Image quality produced in radiotherapy simulation is often so poor in certain anatomical sites (e.g. the pelvis) that it can adversely affect clinical judgement with regard to treatment set-up. In treatment planning both the field arrangement and the availability of accurate anatomical information are important, and should ideally be provided at the time of simulation by methods based on the principle of computerised tomography. In this work an image processing system has been used as a link between the video output from the simulator and the treatment planning computer in order to implement a CT facility on the simulator, as well as allowing real-time image enhancement to be investigated as a method of improving picture quality. This approach also enables a new method of simulation to be considered which reduces patient dose and prolongs the useful working life of the X-ray tube.

Two hundred and twenty-one patients with invasive bladder cancer had routine bone scintigraphy performed as part of tumour staging. The incidence of detectable metastases was 12%. Skeletal scintigraphy at diagnosis has a sensitivity of only 38%. The predictive value of a negative examination was 92%. These results suggest that skeletal scintigraphy is not indicated as a routine staging procedure in carcinoma of the bladder.