Search

The purpose of this work is to provide guidelines for the routine use of portal dosimetry and in vivo diode measurements to verify intensity-modulated radiotherapy (IMRT) treatments. To achieve tolerance levels that are sensitive enough to intercept problems, both the portal dosimetry and the in vivo procedure must be optimised. Portal dosimetry was improved by the introduction of an optimised twodimensional (2D) profile correction, which also accounted for the effect of backscatter from the R-arm. The scaled score, indicating the fraction of points not meeting the desired gamma evaluation criteria within the field opening, was determined as the parameter of interest. Using gamma criteria of a 3%dose difference and 3 mm distance to agreement, a ''scaled score'' threshold value of 1.5% was chosen to indicate excessive tongue and groove and other problems. The pre-treatment portal dosimetry quality assurance (QA) does not encompass verification of the patient dose calculation or position, and so it is complemented by in vivo diode measurements. Diode positioning is crucial in IMRT, and so we describe a method for diode positioning at any suitable point. We achieved 95% of IMRT field measurements within ±5% and 99% within ±8%, with improved accuracy being achieved over time owing to better positioning. Although the careful preparation and setup of the diode measurements can be time-consuming, this is compensated for by the time efficiency of the optimised procedure. Both methods are now easily absorbed into the routine work of the department. [ABSTRACT FROM AUTHOR]

Background and purpose: Due to the increased complexity of the radiotherapy process in recent years, many QA procedures have been introduced to improve the quality of the treatment preparation and the treatment delivery. Most of these procedures give no direct information on the efficiency of the overall process in the radiotherapy department. It is quite evident that the frequency of errors will affect the efficiency of the process (for instance manpower to solve the detected problems), but also affect more generally the patients' experience.Material and methods: To make some objective assessment of the erroneous entries in the process of radiotherapy, a pilot study was initiated to report all problems during a period of 2 months on a dedicated QA form. By registration of these errors, the QA group could trace the main sources of the decrease in efficiency and increased waiting time for patients.Results and conclusions: For 46% of the patients during the two months at least one QA forms was filled in. With these QA forms 198 erroneous entries including 33 erroneous treatments were reported. Regarding the origin of the problems, the majority (60%) were generated in the planning process, while 56% of the errors were detected at the treatment unit before or during delivery of the first fraction. In total 461 actions were undertaken to recover all the reported errors. This generated time delay for 145 patients. Extrapolating these results to a year, a 0.7 full time equivalent person is spent to recover these errors, while the total patient population waits for about 140 hours. Corrective actions have been undertaken to optimise the overall process. [ABSTRACT FROM PUBLISHER]

Automatic segmentation of anatomical structures in medical images is a valuable tool for efficient computer-aided radiotherapy and surgery planning and an enabling technology for dynamic adaptive radiotherapy. This paper presents the design, algorithms and validation of new software for the automatic segmentation of CT images used for radiotherapy treatment planning. A coarse to fine approach is followed that consists of presegmentation, anatomic orientation and structure segmentation. No user input or a priori information about the image content is required. In presegmentation, the body outline, the bones and lung equivalent tissue are detected. Anatomic orientation recognizes the patient's position, orientation and gender and creates an elastic mapping of the slice positions to a reference scale. Structure segmentation is divided into localization, outlining and refinement, performed by procedures with implicit anatomic knowledge using standard image processing operations. The presented version of algorithms automatically segments the body outline and bones in any gender and patient position, the prostate, bladder and femoral heads for male pelvis in supine position, and the spinal canal, lungs, heart and trachea in supine position. The software was developed and tested on a collection of over 600 clinical radiotherapy planning CT stacks. In a qualitative validation on this test collection, anatomic orientation correctly detected gender, patient position and body region in 98% of the cases, a correct mapping was produced for 89% of thorax and 94% of pelvis cases. The average processing time for the entire segmentation of a CT stack was less than 1 min on a standard personal computer. Two independent retrospective studies were carried out for clinical validation. Study I was performed on 66 cases (30 pelvis, 36 thorax) with dosimetrists, study II on 52 cases (39 pelvis, 13 thorax) with radio-oncologists as experts. The experts rated the automatically produced structures on the scale 1-excellent (no corrections necessary, maximum time saving), 2-good (corrections necessary for up to 1/3 of slices), 3-acceptable (major corrections necessary, but still time saving), 4-not acceptable (manual redrawing more efficient, no time saving). A rating

Background and Purpose: This study presents a technical description of the solid multipurpose phantom (MPP) 'OPERA' (operational phantom for external radiotherapy audit) for mailed dosimetry checks of photon beams in reference and non-reference conditions. The aim is to test the functionality of this phantom by comparing the results provided by the local treatment planning system (TPS) with the measurements by film and thermoluminescence dosimeters (TLDs) in a number of clinical conditions., Material and Methods: The polystyrene MPP is constructed to check the following five irradiation conditions: square fields, asymmetrical fields, wedged beams, oblique incidence and influence of inhomogeneities in the field. The absorbed dose on the central beam axis is measured with TLDs for the first three irradiation conditions and the relative dose distributions are verified with film., Results: The 'OPERA' phantom and the corresponding instruction sheets were mailed to radiotherapy centres of 12 different countries, in order to verify its functionality. A good agreement between TPS and the film data is found: the mean deviations between relative dose profiles obtained from film and locally applied TPS data range between -1.2 and 0.4%, except for the short side of the asymmetrical field. The standard deviations (SDs) generally increase off-axis, with a large SD on the asymmetrical side. Considering the TLD results, a mean ratio of measured to stated dose of 0.995+/-0.014 (1 SD), 0.988+/-0.019 (1 SD) and 0.981+/-0.024 (1 SD) for set-up 1 (square field), 2 (asymmetrical field) and 3 (wedged field), respectively, has been found., Conclusion: The 'OPERA' phantom can be useful for on- and off-axis verification of the TPS, as well as for mailed dosimetrical checks of photon beams in reference and non-reference conditions.

Background and Purpose: Quality assurance (QA) in intensity modulated treatments is a complex and time-consuming process. In spite of intensive quality control procedures some types of errors still can go undetected through the course of a treatment. This study aims to develop an objective QA filter for fast, automatic detection of errors, based on the creation of a global platform monitoring treatment parameters by comparison with existing local standards. Since such a conceptually new type of QA has already proven successful for conventional treatments, we aim to extend it to intensity modulated radiotherapy (IMRT) treatments., Material and Methods: The feasibility of developing such a QA platform was evaluated on 12 prostate treatment plans. Apart from the classical treatment parameters, the optimised fluence distributions were compared and screened for repetitive patterns. This screening was performed by calculating specific parameters, defined to characterise the fluence maps., Results: Repetitive patterns were found in the beam shapes as well as in the intensity distributions, and useful parameters could be defined to quantify typical field patterns. By statistical analysis of these parameters, mean values and tolerance levels were derived, providing a means to automatically filter out unprobable or erroneous intensity modulated treatment fields during the treatment preparation stage., Conclusion: It is possible to identify parameters quantifying the characteristic patterns found in fluence distributions of intensity modulated fields of a specific treatment, allowing the development of a platform for automatic pre-treatment quality control.

Background and Purpose: Although intensity modulated radiation therapy is characterized by three-dimensional dose distributions which are often superior to those obtained with conventional treatment plans, its routine clinical implementation is partially held back by the complexity of the beam verification. This is even more so when a dynamic multileaf collimator (dMLC) is used instead of a segmented beam delivery. We have therefore investigated the possibility of using a commercially available, liquid-filled electronic portal imaging device (EPID) for the pre-treatment quality assurance of dynamically delivered dose distributions., Methods and Materials: A special acquisition mode was developed to optimize the image acquisition speed for dosimetry with the liquid-filled EPID. We investigated the accuracy of this mode for 6 and 18 MV photon beams through comparison with film and ion chamber measurements. The impact of leaf speed and pulse rate fluctuations was quantified by means of dMLC plans especially designed for this purpose. Other factors influencing the accuracy of the dosimetry (e.g. the need for build-up, remanence of the ion concentration in the liquid and bulging of the liquid at non-zero gantry angles) were studied as well. We finally compared dosimetric EPID images with the corresponding image prediction delivered without a patient in the beam., Results: The dosimetric accuracy of the measured dose distribution is approximately 2% with respect to film and ion chamber measurements. The accuracy declines when leaf speed is increased beyond 2 cm/s, but is fairly insensitive to accelerator pulse rate fluctuations. The memory effect is found to be of no clinical relevance. When comparing the acquired and expected distributions, an overall agreement of 3% can be obtained, except at areas of steep dose gradients where slight positional shifts are translated into large errors., Conclusions: Accurate dosimetric images of intensity modulated beam profiles delivered with a dMLC can be obtained with a commercially available, liquid-filled EPID. The developed acquisition mode is especially suited for fast and accurate pre-treatment verification of the intensity modulated fields.

Purpose: To present the results of the dummy run of the European Organization for Research and Treatment of Cancer (EORTC) trial investigating the role of adjuvant internal mammary and medial supraclavicular (IM-MS) irradiation in Stage I--III breast cancer., Methods and Materials: All participating institutions were asked to produce a treatment plan without (Arm 1) and with (Arm 2) simultaneous IM-MS irradiation of 1 patient after mastectomy and of 1 patient after lumpectomy. Thirty-two dummy runs have been evaluated for compliance to protocol guidelines, with respect to treatment technique and dose prescription., Results: A number of more or less important deviations in treatment setup and prescription have been found. The dose in the IM-MS region deviated significantly from the prescribed dose in 10% of the cases for Arm 1, and in 21% for Arm 2. Assuming a true 5% 10-year survival benefit from optimal IM-MS irradiation, an increase of only 3.8% will be found due to this suboptimal dose distribution., Conclusion: In the dummy run, a number of potential systematic protocol deviations that might lead to false-negative results were detected. By providing recommendations to the participating institutions, we expect to improve the interinstitutional consistency and to promote a high quality irradiation in all institutions participating in the trial.

Aim: A feasibility study has been performed to investigate the possibility of using mailed thermoluminescence dosimetry (TLD) for external audits of clinical electron beams in Europe., Methods: In the frame of the EC Network Project for Quality Assurance in Radiotherapy, instruction sheets and mailing procedures have been defined for mailed TLD dosimetry using the dedicated holder developed by a panel of experts of the International Atomic Energy Agency (IAEA). Three hundred and thirty electron beam set-ups have been checked in the reference centres and some local centres of the EC Network Project and in addition through the centres participating to the EORTC Radiotherapy Group trial 22922., Results: The mean ratio of measured dose to stated dose is 0.2% and the standard deviation of measured dose to stated dose is 3.2%. In seven beam set-ups, deviations greater than 10% were observed (max. 66%), showing the usefulness of these checks., Conclusion: The results of this feasibility study (instruction sheets, mailing procedures, holder) are presently endorsed by the EQUAL-ESTRO structure in order to offer in the future to all ESTRO members the possibility to request external audits of clinical electron beams.

Background and Purpose: In the Radiotherapy Department of Leuven, about 20% of all breast cancer patients treated with breast conserving surgery and external radiotherapy receive an additional boost with pulsed dose rate (PDR) Ir-192 brachytherapy. An investigation was performed to assess the accuracy of the delivered PDR brachytherapy treatment. Secondly, the feasibility of in vivo measurements during PDR dose delivery was investigated., Materials and Methods: Two phantoms are manufactured to mimic a breast, one for thermoluminescent dosimetry (TLD) measurements, and one for dosimetry using radiochromic films. The TLD phantom allows measurements at 34 dose points in three planes including the basal dose points. The film phantom is designed in such a way that films can be positioned in a plane parallel and orthogonal to the needles., Results: The dose distributions calculated with the TPS are in good agreement with both TLD and radiochromic film measurements (average deviations of point doses <+/-5%). However, close to the interface tissue-air the dose is overestimated by the TPS since it neglects the finite size of a breast and the associated lack of backscatter (average deviations of point doses -14%)., Conclusion: Most deviations between measured and calculated doses, are in the order of magnitude of the uncertainty associated with the source strength specification, except for the point doses measured close to the skin. In vivo dosimetry during PDR brachytherapy treatment was found to be a valuable procedure to detect large errors, e.g. errors caused by an incorrect data transfer.

Background and Purpose: Due to the complexity of the treatment preparation in radiotherapy, a number of errors go undetected until after the first treatment session. Some of these errors could easily have been noticed before treatment if an objective filter existed in addition to human supervision. With this in mind, a conceptually novel extension to conventional quality assurance procedures was explored to create a global platform monitoring treatment preparation by comparison with the existing local standards., Materials and Methods: The feasibility of developing such a platform was evaluated for a test case on a cohort of 202 patients having received breast irradiation. By statistical analysis of the treatment parameters, mean values and tolerance levels could be defined for most parameters based on the observed standard deviations. Useful correlations were traced providing us with a means to automatically track errors, the detection of which would otherwise solely depend upon the alertness of the supervisor., Results and Conclusions: Apart from its possibilities as a mere quality control tool, the platform, developed in the framework of EQUART (European Quality Assurance Program in Radiotherapy by Monitoring Treatment Preparation), can be incorporated in the treatment preparation chain, providing standard setup values for the simulation. A crucial achievement of EQUART lies in the fact that filtering out of errors occurs prior to treatment initiation.

Purpose: (18)F-fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET) combined with computer tomography (PET-CT) is superior to CT alone in mediastinal lymph node (LN) staging in non-small cell lung cancer (NSCLC). We studied the potential impact of this non-invasive LN staging procedure on the radiation treatment plan of patients with NSCLC., Patients and Methods: The imaging and surgical pathology data from 105 patients included in two previously published prospective LN staging protocols form the basis for the present analysis. For 73 of these patients, with positive LN's on CT and/or on PET, a theoretical study was performed in which for each patient the gross tumour volume (GTV) was defined based on CT and on PET-CT data. For each GTV, the completeness of tumour coverage was assessed, using the available surgical pathology data as gold standard. A more detailed analysis was done for the first ten consecutive patients in whom the PET-CT-GTV was smaller than the CT-GTV. Theoretical radiation treatment plans were constructed based on both CT-GTV and PET-CT-GTV. Dose-volume histograms for the planning target volume (PTV), for the total lung volume and the lung volume receiving more than 20 Gy (V(lung(20))), were calculated., Results: Data from 988 assessed LN stations were available. In the subgroup of 73 patients with CT or PET positive LN's, tumour coverage improved from 75% when the CT-GTV was used to 89% with the PET-CT-GTV (P=0.005). In 45 patients (62%) the information obtained from PET would have led to a change of the treatment volumes. For the ten patients in the dosimetry study, the use of PET-CT to define the GTV, resulted in an average reduction of the PTV by 29+/-18% (+/-1 SD) (P=0.002) and of the V(lung(20)) of 27+/-18% (+/-1 SD) (P=0.001)., Conclusion: In patients with NSCLC considered for curative radiation treatment, assessment of locoregional LN tumour extension by PET will improve tumour coverage, and in selected patients, will reduce the volume of normal tissues irradiated, and thus toxicity. This subgroup of patients could then become candidates for treatment intensification.

Aim: To present a technical description and intercomparison of three multipurpose phantoms (MPP) developed for mailed dosimetry checks of therapeutic photon beams in reference and non-reference conditions., Materials: The W-MPP is a water MPP, whereas the Umeâ-MPP, made of perspex (PMMA, Plexiglas), and the EC-MPP, made of polystyrene, are solid MPPs. The W-MPP uses only thermoluminescent dosimeters (TLD) for dosimetry checks, the EC MPP uses film and TLD; the Umeâ phantom uses film and TLD, and offers in addition the possibility for ionization chamber measurements. Either using TLD or films, the MPPs have been designed to check on-axis and off-axis the following irradiation conditions: square and rectangular fields, asymmetric fields, wedged beams, oblique incidence and, for the solid MPPs, also the influence of inhomogeneities., Results and Discussion: The MPPs have been compared for different aspects going from their dosimetric performance (number of dosimetric parameters that can be checked) to some practical consideration in the use of the different MPPs (set-up time, stability, instruction sheets, etc.). From a comparison between the solid multi-purpose phantoms, it turns out that the EC-MPP is capable of checking the largest number of dosimetric parameters per beam, but has the longest set-up time ( approximately 2 h) per beam according to the users. The Umeå-MPP has a smaller number of set-ups (hence a smaller average time) and also includes some parameters not checked with the EC-MPP (e.g. SSD accuracy). The major drawback however of the Umeå-MPP is considered to be its high density (>1.1 g/cm(3)) which increases the difficulty of the analysis with the treatment planning system. The W-MPP checks the smallest number of parameters, but is the fastest in set-up time, the easiest for mailing, and is water equivalent, which is advantageous for the TPS checks. The major drawback of this MPP is however the inability to check complete dose distribution (film) or inhomogeneities.

Purpose and Background: A feasibility study is performed to evaluate the possibility of using the transit dose of portal images on the beam axis to measure the accuracy in dose delivery. The algorithm and the method are tested on a breast phantom and on patients with a breast disease., Materials and Methods: To estimate the transit dose at various air gaps behind the patient, a method is proposed which applies, for a given air gap, the inverse square law to the primary component of the exit dose and an experimentally determined function for the scatter component of the exit dose. It is assumed that the primary component and the scattered component of the exit dose are given by the treatment planning system. The experimental function for the variation of the scattered component with the air gap, determined by phantom measurements, is modelled by an analytical function which contains only field size, air gap and one energy-dependent parameter., Results: The measurements on the breast phantom yield a maximum deviation between measured and estimated transit doses of 4.5%. The mean deviation is 0.9% with a standard deviation of the distribution of 2.3%. In vivo diode measurements on the same phantom yield a maximum deviation of 2.7%. Transit dose measurements on the beam axis for 45 portal images of breast patients show a mean deviation of 0.0% between the measured transit dose and the estimated transit dose. The standard deviation of the distribution is 4.4%. The method seems to be very sensitive to patient positioning and to discrepancies in breast thicknesses used for treatment planning., Conclusion: Preliminary results on breast patients show that the method proposed to evaluate transit doses on the beam axis from portal images may be a valuable alternative to conventional in vivo exit dosimetry. The method can be implemented in a simple way and does not require additional time during the irradiation session, as exit dosimetry with diodes does. The transit dose is only considered in one point. Nevertheless, in the framework of quality assurance of treatment delivery, this study is an example of the possibilities of monitoring at the same time the visual evaluation of the irradiated volume as well as the dosimetric control (i.e. in Gy) of treatment delivery with portal images.

Background and Purpose: In view of using portal images for exit dosimetry, an experimental study is performed of relative transit dose profiles at different distances behind patients (and phantoms) and of their relation to the exit dose profile., Materials and Methods: Irregular, homogeneous polystyrene phantoms with a variable thickness to simulate head and neck (H&N) treatments (6-MV photon beam) are investigated by ionization chamber measurements performed close to the exit surface and at various distances behind the phantom (10, 20 and 30 cm). Similar measurements are performed for a rectangular phantom with large inhomogeneities (A1 and air). For one irregular homogeneous phantom and an irregular phantom containing an A1 inhomogeneity, ionization chamber measurements are performed at the exit surface, and a portal film image is taken at 30 cm behind the phantom. Portal films of a patient treated for a head and neck malignancy are evaluated for different air gaps behind the patient., Results: For the irregular phantoms, deviations up to 15% and more are observed between the exit dose profile (along the shaped surface of the phantom) and the transit profile close to the phantom (perpendicular to the beam axis). There is, however, a good agreement--within 3%--between the exit profile and the transit profile at 30 cm. For the rectangular, inhomogeneous phantom, the deviation between the exit profile and the transit dose profile at 30 cm does not exceed 5%; transit dose profiles overestimate the exit dose for the air cavity and underestimate the dose for the A1 inhomogeneity. Measurements on portal films of a H&N patient for different air gaps confirm the order of magnitude of the difference observed between transit dose profiles close to the patient and transit dose profiles at some distance behind the patient., Conclusions: For 6-MV photon beam treatments with significant thickness variations (H&N), large variations (> 10%) are observed in transit dose profiles as a function of the air gap between the patient and the portal film. For this energy, a good agreement is found between the exit profile and the transit profile at about 30 cm behind the patient.

Background and Purpose: Several commercially available p-type diodes do not provide sufficient build-up for in-vivo dosimetry in 'higher' energy photon beams, and only limited information could be found in the literature describing the correction factor variation and/or the achievable accuracy for in-vivo dosimetry methods in this energy range. The first aim of this study is to assess and analyze the variation of diode correction factors for entrance dose measurements at higher photon energies. In a second step the total build up thickness of the diode has been modified in order to minimize the correction factor variation., Materials and Methods: Diode correction factors accounting for non-reference conditions (field size, source surface distance, tray, wedge, and block) are determined in 18-25 MV photon beams provided by different treatment units for Scanditronix p-type diodes recommended for higher energy photon beams: old type and new type EDP-20, and EDP-30 diodes. Hemispherical build-up caps of different materials (copper, iron, lead) are used to increase the total build-up thickness. Perturbation effects with and without additional build-up caps are assessed for the three diode types., Results: For unmodified diodes field size correction factors (C(FS)) vary between 1.7% and 6%, dependent on diode type and treatment unit. For example, for an old type EDP-20 the C(FS) variation at 18 MV is much higher on a GE linac (5%) as compared to the Philips machine (1.7%). Depending on diode type, this variation can be reduced to 1-2% when adding additional build-up. The variation of source to surface distance correction factors is almost independent of build-up thickness. By adding additional build-up the influence of trays and blocks can be almost eliminated., Conclusions: The correction factor variation of unmodified diodes reflects the variation of the electron contamination with treatment geometry. A total build-up thickness of 30 mm is found to be the 'best compromise' for the three types of diodes investigated when measuring entrance doses in the energy range between 18 and 25 MV.

The possibility of using a commercially available electronic portal imaging device for transit dosimetry was investigated. The detection unit of the device comprises a metal plate/fluorescence screen and a camera. Basic parameters of this system were investigated: stability, detector uniformity, dose-response curve, field-size dependence and phantom-thickness dependence. It was found that in terms of relative dosimetry, portal images corrected for detector non-uniformity are in good agreement (within 5%) with transit dose distributions measured by film dosimetry. For dose determination, it was found that the use of the device is hampered by an important field-size dependence and phantom-thickness dependence. Both correction factors should be applied if the device is to be used for this purpose.

Aim: To use portal images acquired in routine circumstances for assessment of midplane dose variations in the patient., Material and Methods: Optical density readings are performed on routinely acquired Verification films of breast and ear-nose-throat (ENT) cancer patients and these readings are converted into relative doses with the sensitometric curve. ( 1 ) The impact of redistribution is evaluated on films taken close to the patient exit surface and at routine focus film distance. (2) Midplane doses are estimated from film readings to assess dose variations in the patient. The influence of wedges is evaluated. Film measurements doses are compared with calculated exit doses., Results: (1) In regions with large variations in the distance between the patient exit surface and the film but without inhomogeneities in tissue density, the relative doses distributions read on films acquired at large focus-film-distance (FFD) are proportional to exit doses. In regions with flat exit surfaces but with inhomogeneities in tissue density, the redistribution has only a small impact. (2) Large variations in relative midplane doses were found in both breast (85-115%) and ENT (-3.6 to +15%) patients. The application of a wedge was shown to increase dose homogeneity in the midplane. A good agreement (differences < 3%) was found between exit doses obtained from film readings and exit doses calculated with the treatment planning system (TPS)., Conclusion: Films acquired in routine circumstances at large FFD can be used to obtain information on exit doses and to assess midplane doses in breast and ENT, without the use of a TPS. Film dosimetry can also provide a quality assurance tool to check actually delivered doses in patients by comparing exit doses estimated on film to expected exit doses calculated by the TPS.