Artificial intelligence in radiation oncology: A review of its current status and potential application for the radiotherapy workforce.

OBJECTIVE: Radiation oncology is a continually evolving speciality. With the development of new imaging modalities and advanced imaging processing techniques, there is an increasing amount of data available to practitioners. In this narrative review, Artificial Intelligence (AI) is used as a reference to machine learning, and its potential, along with current problems in the field of radiation oncology, are considered from a technical position. KEY FINDINGS: AI has the potential to harness the availability of data for improving patient outcomes, reducing toxicity, and easing clinical burdens. However, problems including the requirement of complexity of data, undefined core outcomes and limited generalisability are apparent. CONCLUSION: This original review highlights considerations for the radiotherapy workforce, particularly therapeutic radiographers, as there will be an increasing requirement for their familiarity with AI due to their unique position as the interface between imaging technology and patients. IMPLICATIONS FOR PRACTICE: Collaboration between AI experts and the radiotherapy workforce are required to overcome current issues before clinical adoption. The development of educational resources and standardised reporting of AI studies may help facilitate this.

Head and neck target delineation using a novel PET automatic segmentation algorithm.

PURPOSE: To evaluate the feasibility and impact of using a novel advanced PET auto-segmentation method in Head and Neck (H&N) radiotherapy treatment (RT) planning. METHODS: ATLAAS, Automatic decision Tree-based Learning Algorithm for Advanced Segmentation, previously developed and validated on pre-clinical data, was applied to 18F-FDG-PET/CT scans of 20 H&N patients undergoing Intensity Modulated Radiation Therapy. Primary Gross Tumour Volumes (GTVs) manually delineated on CT/MRI scans (GTVpCT/MRI), together with ATLAAS-generated contours (GTVpATLAAS) were used to derive the RT planning GTV (GTVpfinal). ATLAAS outlines were compared to CT/MRI and final GTVs qualitatively and quantitatively using a conformity metric. RESULTS: The ATLAAS contours were found to be reliable and useful. The volume of GTVpATLAAS was smaller than GTVpCT/MRI in 70% of the cases, with an average conformity index of 0.70. The information provided by ATLAAS was used to grow the GTVpCT/MRI in 10 cases (up to 10.6mL) and to shrink the GTVpCT/MRI in 7 cases (up to 12.3mL). ATLAAS provided complementary information to CT/MRI and GTVpATLAAS contributed to up to 33% of the final GTV volume across the patient cohort. CONCLUSIONS: ATLAAS can deliver operator independent PET segmentation to augment clinical outlining using CT and MRI and could have utility in future clinical studies.

Superiority of Deformable Image Co-registration in the Integration of Diagnostic Positron Emission Tomography-Computed Tomography to the Radiotherapy Treatment Planning Pathway for Oesophageal Carcinoma.

AIMS: To investigate the use of image co-registration in incorporating diagnostic positron emission tomography-computed tomography (PET-CT) directly into the radiotherapy treatment planning pathway, and to describe the pattern of local recurrence relative to the PET-avid volume. MATERIALS AND METHODS: Fourteen patients were retrospectively identified, six of whom had local recurrence. The accuracy of deformable image registration (DIR) and rigid registration of the diagnostic PET-CT and recurrence CT, to the planning CT, were quantitatively assessed by comparing co-registration of oesophagus, trachea and aorta contours. DIR was used to examine the correlation between PET-avid volumes, dosimetry and site of recurrence. RESULTS: Positional metrics including the dice similarity coefficient (DSC) and conformity index (CI), showed DIR to be superior to rigid registration in the co-registration of diagnostic and recurrence imaging to the planning CT. For diagnostic PET-CT, DIR was superior to rigid registration in the transfer of oesophagus (DSC=0.75 versus 0.65, P<0.009 and CI=0.59 versus 0.48, P<0.003), trachea (DSC=0.88 versus 0.65, P<0.004 and CI=0.78 versus 0.51, P<0.0001) and aorta structures (DSC=0.93 versus 0.86, P<0.006 and CI=0.86 versus 0.76, P<0.006). For recurrence imaging, DIR was superior to rigid registration in the transfer of trachea (DSC=0.91 versus 0.66, P<0.03 and CI=0.83 versus 0.51, P<0.02) and oesophagus structures (DSC=0.74 versus 0.51, P<0.004 and CI=0.61 versus 0.37, P<0.006) with a non-significant trend for the aorta (DSC=0.91 versus 0.75, P<0.08 and CI=0.83 versus 0.63, P<0.06) structure. A mean inclusivity index of 0.93 (range 0.79-1) showed that the relapse volume was within the planning target volume (PTVPET-CT); all relapses occurred within the high dose region. CONCLUSION: DIR is superior to rigid registration in the co-registration of PET-CT and recurrence CT to the planning CT, and can be considered in the direct integration of PET-CT to the treatment planning process. Local recurrences occur within the PTVPET-CT, suggesting that this is a suitable target for dose-escalation strategies.

A novel phantom technique for evaluating the performance of PET auto-segmentation methods in delineating heterogeneous and irregular lesions.

BACKGROUND: Positron Emission Tomography (PET)-based automatic segmentation (PET-AS) methods can improve tumour delineation for radiotherapy treatment planning, particularly for Head and Neck (H&N) cancer. Thorough validation of PET-AS on relevant data is currently needed. Printed subresolution sandwich (SS) phantoms allow modelling heterogeneous and irregular tracer uptake, while providing reference uptake data. This work aimed to demonstrate the usefulness of the printed SS phantom technique in recreating complex realistic H&N radiotracer uptake for evaluating several PET-AS methods. METHODS: Ten SS phantoms were built from printouts representing 2mm-spaced slices of modelled H&N uptake, printed using black ink mixed with 18F-fluorodeoxyglucose, and stacked between 2mm thick plastic sheets. Spherical lesions were modelled for two contrasted uptake levels, and irregular and spheroidal tumours were modelled for homogeneous, and heterogeneous uptake including necrotic patterns. The PET scans acquired were segmented with ten custom PET-AS methods: adaptive iterative thresholding (AT), region growing, clustering applied to 2 to 8 clusters, and watershed transform-based segmentation. The difference between the resulting contours and the ground truth from the image template was evaluated using the Dice Similarity Coefficient (DSC), Sensitivity and Positive Predictive value. RESULTS: Realistic H&N images were obtained within 90 min of preparation. The sensitivity of binary PET-AS and clustering using small numbers of clusters dropped for highly heterogeneous spheres. The accuracy of PET-AS methods dropped between 4% and 68% for irregular lesions compared to spheres of the same volume. For each geometry and uptake modelled with the SS phantoms, we report the number of clusters resulting in optimal segmentation. Radioisotope distributions representing necrotic uptakes proved most challenging for most methods. Two PET-AS methods did not include the necrotic region in the segmented volume. CONCLUSIONS: Printed SS phantoms allowed identifying advantages and drawbacks of the different methods, determining the most robust PET-AS for the segmentation of heterogeneities and complex geometries, and quantifying differences across methods in the delineation of necrotic lesions. The printed SS phantom technique provides key advantages in the development and evaluation of PET segmentation methods and has a future in the field of radioisotope imaging.

Evaluation of advanced automatic PET segmentation methods using nonspherical thin-wall inserts.

PURPOSE: The use of positron emission tomography (PET) within radiotherapy treatment planning requires the availability of reliable and accurate segmentation tools. PET automatic segmentation (PET-AS) methods have been recommended for the delineation of tumors, but there is still a lack of thorough validation and cross-comparison of such methods using clinically relevant data. In particular, studies validating PET segmentation tools mainly use phantoms with thick plastic walls inserts of simple spherical geometry and have not specifically investigated the effect of the target object geometry on the delineation accuracy. Our work therefore aimed at generating clinically realistic data using nonspherical thin-wall plastic inserts, for the evaluation and comparison of a set of eight promising PET-AS approaches. METHODS: Sixteen nonspherical inserts were manufactured with a plastic wall of 0.18 mm and scanned within a custom plastic phantom. These included ellipsoids and toroids derived with different volumes, as well as tubes, pear- and drop-shaped inserts with different aspect ratios. A set of six spheres of volumes ranging from 0.5 to 102 ml was used for a baseline study. A selection of eight PET-AS methods, written in house, was applied to the images obtained. The methods represented promising segmentation approaches such as adaptive iterative thresholding, region-growing, clustering and gradient-based schemes. The delineation accuracy was measured in terms of overlap with the computed tomography reference contour, using the dice similarity coefficient (DSC), and error in dimensions. RESULTS: The delineation accuracy was lower for nonspherical inserts than for spheres of the same volume in 88% cases. Slice-by-slice gradient-based methods, showed particularly lower DSC for tori (DSC < 0.5), caused by a failure to recover the object geometry. The region-growing method reached high levels of accuracy for most inserts (DSC > 0.76 except for tori) but showed the largest errors in the recovery of pears and drops dimensions (higher than 10% and 30% of the true length, respectively). Large errors were visible for one of the gradient-based contouring methods when delineating drop-shaped inserts. Low DSC due to systematic underestimation of the volumes was observed for our fuzzy clustering method when using nonspherical inserts. The adaptive iterative thresholding method produced the highest DSC score on our nonspherical dataset (DSC > 0.83, except for tori) and showed robustness to the insert geometry. CONCLUSIONS: This study investigated the accuracy of eight PET-AS methods for the delineation of objects with a range of nonspherical geometries. The authors' results confirmed the robustness of some segmentation approaches, but also showed the weaknesses of some of the other methods implemented, which were not observed with spherical inserts. This work therefore highlights the importance of using a variety of thin-wall inserts with complex geometries for the validation of PET-AS methods and provided useful information for further development of the methods tested.

Influence of cold walls on PET image quantification and volume segmentation: a phantom study.

PURPOSE: Commercially available fillable plastic inserts used in positron emission tomography phantoms usually have thick plastic walls, separating their content from the background activity. These "cold" walls can modify the intensity values of neighboring active regions due to the partial volume effect, resulting in errors in the estimation of standardized uptake values. Numerous papers suggest that this is an issue for phantom work simulating tumor tissue, quality control, and calibration work. This study aims to investigate the influence of the cold plastic wall thickness on the quantification of 18F-fluorodeoxyglucose on the image activity recovery and on the performance of advanced automatic segmentation algorithms for the delineation of active regions delimited by plastic walls. METHODS: A commercial set of six spheres of different diameters was replicated using a manufacturing technique which achieves a reduction in plastic walls thickness of up to 90%, while keeping the same internal volume. Both sets of thin- and thick-wall inserts were imaged simultaneously in a custom phantom for six different tumor-to-background ratios. Intensity values were compared in terms of mean and maximum standardized uptake values (SUVs) in the spheres and mean SUV of the hottest 1 ml region (SUVmax, SUVmean, and SUVpeak). The recovery coefficient (RC) was also derived for each sphere. The results were compared against the values predicted by a theoretical model of the PET-intensity profiles for the same tumor-to-background ratios (TBRs), sphere sizes, and wall thicknesses. In addition, ten automatic segmentation methods, written in house, were applied to both thin- and thick-wall inserts. The contours obtained were compared to computed tomography derived gold standard ("ground truth"), using five different accuracy metrics. RESULTS: The authors' results showed that thin-wall inserts achieved significantly higher SUVmean, SUVmax, and RC values (up to 25%, 16%, and 25% higher, respectively) compared to thick-wall inserts, which was in agreement with the theory. This effect decreased with increasing sphere size and TBR, and resulted in substantial (>5%) differences between thin- and thick-wall inserts for spheres up to 30 mm diameter and TBR up to 4. Thinner plastic walls were also shown to significantly improve the delineation accuracy for the majority of the segmentation methods tested, by increasing the proportion of lesion voxels detected, although the errors in image quantification remained non-negligible. CONCLUSIONS: This study quantified the significant effect of a 90% reduction in the thickness of insert walls on SUV quantification and PET-based boundary detection. Mean SUVs inside the inserts and recovery coefficients were particularly affected by the presence of thick cold walls, as predicted by a theoretical approach. The accuracy of some delineation algorithms was also significantly improved by the introduction of thin wall inserts instead of thick wall inserts. This study demonstrates the risk of errors deriving from the use of cold wall inserts to assess and compare the performance of PET segmentation methods.

Development and validation of RAYDOSE: a Geant4-based application for molecular radiotherapy.

We developed and validated a Monte-Carlo-based application (RAYDOSE) to generate patient-specific 3D dose maps on the basis of pre-treatment imaging studies. A CT DICOM image is used to model patient geometry, while repeated PET scans are employed to assess radionuclide kinetics and distribution at the voxel level. In this work, we describe the structure of this application and present the tests performed to validate it against reference data and experiments. We used the spheres of a NEMA phantom to calculate S values and total doses. The comparison with reference data from OLINDA/EXM showed an agreement within 2% for a sphere size above 2.8 cm diameter. A custom heterogeneous phantom composed of several layers of Perspex and lung equivalent material was used to compare TLD measurements of gamma radiation from (131)I to Monte Carlo simulations. An agreement within 5% was found. RAYDOSE has been validated against reference data and experimental measurements and can be a useful multi-modality platform for treatment planning and research in MRT.

Oesophageal Chemoradiotherapy in the UK--current practice and future directions.

The SCOPE 1 trial closed to recruitment in early 2012 and has demonstrably improved the quality of UK radiotherapy. It has also shown that there is an enthusiastic upper gastrointestinal clinical oncology community that can successfully complete trials and deliver high-quality radiotherapy. Following on from SCOPE 1, this paper, authored by a consensus of leading UK upper gastrointestinal radiotherapy specialists, attempts to define current best practice and the questions to be answered by future clinical studies. The two main roles for chemoradiotherapy (CRT) in the management of potentially curable oesophageal cancer are definitive (dCRT) and neoadjuvant (naCRT). The rates of local failure after dCRT are consistently high, showing the need to evaluate more effective treatments, both in terms of optimal local and systemic therapeutic components. This will be the primary objective of the next planned UK dCRT trial and here we discuss the role of dose escalation and systemic therapeutic options that will form the basis of that trial. The publication of the Dutch 'CROSS' trial of naCRT has shown that this pre-operative approach can both be given safely and offer a significant survival benefit over surgery alone. This has led to the development of the UK NeoSCOPE trial, due to open in 2013. There will be a translational substudy to this trial and currently available data on the role of biomarkers in predicting response to therapy are discussed. Postoperative reporting of the pathology specimen is discussed, with recommendations for the NeoSCOPE trial. Both of these CRT approaches may benefit from recent developments, such as positron emission tomography/computed tomography and four-dimensional computed tomography for target volume delineation, planning techniques such as intensity-modulated radiotherapy and 'type b' algorithms and new treatment verification methods, such as cone-beam computed tomography. These are discussed here and recommendations made for their use.

Improving radiotherapy quality assurance in clinical trials: assessment of target volume delineation of the pre-accrual benchmark case.

As the complexity of radiotherapy (RT) trials increases, issues surrounding target volume delineation will become more important. Some form of outlining assessment prior to trial entry is increasingly being mandated in UK RT trials. This document produced by the Outlining and Imaging Subgroup (OISG) of the National Cancer Research Institute will address methods to reduce interobserver variation in clinical trials and how to conduct an assessment of outlining through a pre-accrual benchmark case. We review currently available methods of describing the variation and identify areas where further work is needed. The OISG would encourage ongoing discussion with chief investigators in order to provide advice on individual aspects of benchmark case assessment for current and future trials.

Usefulness of 124I PET/CT imaging to predict absorbed doses in patients affected by metastatic thyroid cancer and treated with 131I.

AIM: The aim of this work was the evaluation of the usefulness of 124I PET/CT sequential scans to predict absorbed doses to metastatic thyroid cancer in patients undergoing 131I therapy. METHODS: From July 2011 until April 2012 8 patients affected by metastatic thyroid cancer were enrolled. Each patient underwent 4 PET/CT scans at 4, 24, 48, 72 h after the administration of about 74 MBq of 124I. Blood samples and whole body exposure measurements were obtained to calculate blood and red marrow doses. Activity concentrations and lesion volumes obtained from PET/CT images were used to evaluate tumour doses with MIRD formalism and spheres model. The average administered 131I therapeutic activity was 6475 MBq (range: 3700-9250 MBq). RESULTS: 124I PET/CT images showed, with a very good resolution, all 131I avid lesions detected by post therapy whole body scans. The average dose rates for blood, red marrow and lesions were respectively: 6.58E-02 ± 1.64E-02 mGy/MBq, 5.73E-02 ± 1.57E-02 mGy/MBq, 2.22E+01 ± 1.62E+01 mGy/MBq. Three out of eight patients did not show any uptake of 124I in all PET/CT scans, despite high level of TSH and CT detectable lesions. Post-therapy 131I whole body scan confirmed the absence of focal iodine uptake. CONCLUSION: Negative 124I PET/CT images probably could be used as predictive of real absence of iodine avidity, avoiding all toxicity from useless 131I therapy. A higher number of patients is necessary to validate these preliminary results and a project is ongoing to compare MIRD results to voxel dosimetry based on Monte Carlo simulation.

Imaging for target volume delineation in rectal cancer radiotherapy--a systematic review.

The global move towards more conformal radiotherapy for rectal cancer requires better imaging modalities that both visualise the disease accurately and are reproducible; to reduce interobserver variation. This review explores the advances in imaging modalities used in target volume delineation, with a view to make recommendations for current clinical practice and to propose future directions for research. A systematic review was conducted using MEDLINE and EMBASE. Articles considered relevant by the authors were included. Planning with orthogonal films is being replaced by computed tomography (CT) simulation. This is now considered the 'gold standard' and allows conformal three-dimensional planning. Magnetic resonance imaging (MRI) has been shown to overcome some of the limitations of CT and can be used either as a diagnostic image to visually aid planning, or as a 'planning' MRI carried out in the treatment position and co-registered with the planning CT. The latter approach has been shown to change the treated volumes compared with CT and in prostate cancer patients has been shown to reduce interobserver variation. There are remaining issues with four-dimensional motion that are yet to be fully appreciated or overcome. 2-[18F] fluoro-2-deoxy-d-glucose positron emission tomography/CT co-registered with planning CT results in smaller volumes than CT alone and also reduces interobserver variation, but requires further validation before routine implementation. Experimental work utilising novel positron emission tomography tracers and diffusion-weighted MRI shows promise and requires further evaluation. Rigorous quality assurance is important with processing of newer imaging modalities. Further work needs to be conducted into both interobserver variation and the formal evaluation of the clinical benefits of newer imaging modalities. Developments in image-guided radiotherapy are also required to ensure that improvements in target definition at the planning stage are reproducible throughout treatment.

SU-E-J-15: Calculating the Dose from KV Cone Beam CT Within and Outside the Treatment Volume Using a Treatment Planning System.

PURPOSE: To assess the dose from kilovoltage cone beam CT from two image acquisition protocols, pelvis and head and neck, by addition of the dose to the patient treatment plans. METHODS: A total of 20 patients (10 pelvis and 10 head and neck) undergoing radiation therapy were selected and the dose from kV CBCT was calculated using a treatment planning system previously commissioned for this purpose. The imaging dose was added to the CT images used for treatment planning. Daily shifts as a Result of imaging were recorded and applied to imaging beam whenever the sum of the shifts exceeded 0.5 cm. The kV CBCT dose can also be computed prior to planning, in case of IMRT treatments, and used during optimization. RESULTS: The additional dose as a Result of daily CBCT is in the order of few cGy for head and neck and up to 90 cGy for the pelvis cases using the standard head and neck and pelvis protocols. The pelvic dose is especially dependent on patient size, being higher for smaller patients. Due to the low energy of the kV CBCT beam, the maximum energy deposition is at or near the surface with the highest dose being on the patient's left side for the head and neck and on the posterior for the pelvic cases. CONCLUSIONS: Dose from daily kilovoltage CBCT can be added to patient treatment plans using previously commissioned kV CBCT beams in a treatment planning system. In the case of IMRT planning, optimization can be done accounting for kV CBCT dose.

A virtual source model for kilo-voltage cone beam CT: source characteristics and model validation.

PURPOSE: The purpose of this investigation was to study the source characteristics of a clinical kilo-voltage cone beam CT unit and to develop and validate a virtual source model that could be used for treatment planning purposes. METHODS: We used a previously commissioned full Monte Carlo model and new bespoke software to study the source characteristics of a clinical kilo-voltage cone beam CT (CBCT) unit. We identified the main particle sources, their spatial, energy and angular distribution for all the image acquisition presets currently used in our clinical practice. This includes a combination of two energies (100 and 120 kVp), two filters (neutral and bowtie), and eight different x-ray beam apertures. We subsequently built a virtual source model which we validated against full Monte Carlo calculations. RESULTS: We found that the radiation output of the clinical kilo-voltage cone beam CT unit investigated in this study could be reproduced with a virtual model comprising of two sources (target and filtration cone) or three sources (target, filtration cone and bowtie filter) when additional filtration was used. With this model, we accounted for more than 97% of the photons exiting the unit. Each source in our model was characterised by a origin distribution in both X and Y directions, a fluence map, a single energy spectrum for unfiltered beams and a two dimensional energy spectrum for bowtie filtered beams. The percentage dose difference between full Monte Carlo and virtual source model based dose distributions was well within the statistical uncertainty associated with the calculations ( ± 2%, one standard deviation) in all cases studied. CONCLUSIONS: The virtual source that we developed is accurate in calculating the dose delivered from a commercial kilo-voltage cone beam CT unit operating with routine clinical image acquisition settings. Our data have also shown that target, filtration cone, and bowtie filter sources needed to be all included in the model in order to accurately replicate the dose distribution from the clinical radiation beam.

An investigation of the accuracy of Monte Carlo portal dosimetry for verification of IMRT with extended fields.

This work investigated the accuracy of Monte Carlo (MC) simulations of amorphous silicon (a-Si) electronic portal imaging devices (EPIDs) for the dosimetric verification of intensity-modulated radiotherapy (IMRT). In particular, the suitability of the method for verification of head and neck IMRT with extended field segments (approximately 20 cm superior-inferior), covering almost the entire detector area, was studied. A solution involving schematic modelling of backscatter materials has been established to account for non-uniform backscatter to the imager from supporting structures. 96% of points within the IMRT fields evaluated passed a 'gamma' evaluation criterion of 2%, 2 mm at isocentre at a dose rate of 100 MU min(-1) with this solution included. Only 79% of points passed this gamma criterion without the correction for backscatter included. This work has also demonstrated the ability of the technique to detect systematic delivery errors in step and shoot IMRT. The technique identified a systematic overshoot on the first segment and an undershoot on the final segment. Results were verified by ion chamber measurements and agreed well with those reported in the literature, averaging approximately 0.1 and 0.3 MU for 100 and 300 MU min(-1) deliveries, respectively. MC portal verification has the potential to become a key tool in the verification of IMRT and can also facilitate selection of optimal delivery parameters, thus improving treatment accuracy. This approach can be applied to the verification of other new treatment techniques and should also enable development of methodologies to detect and correct for delivery errors, both before and during treatment.

Monte Carlo simulation and patient dosimetry for a kilovoltage cone-beam CT unit.

PURPOSE: The purpose of this work is to characterize the x-ray volume imager (XVI), the cone-beam computed tomography (CBCT) unit mounted on the Elekta Synergy linac, with F1 bowtie filter and to calculate the three-dimensional dose delivered to patients using volumetric acquisition. METHODS: The XVI is modeled in detail using a new Monte Carlo (MC) code, BEAMPP, under development at the National Research Council Canada. In this investigation, a new component module is developed to accurately model the unit's bowtie filter used in conjunction with the available beam collimators at the clinical energy of 120 kV. The modeling is compared against percentage depth dose (PDD) and profile measurements. Kilovoltage radiation beams' phase space files are also analyzed. The authors also describe a method for the absolute dose calibration of the MC model of the CBCT unit when used in a clinical volumetric acquisition mode. Finally, they calculate three-dimensional patient dose from CBCT image acquisition in three clinical cases of interest: Pelvis, lung, and head and neck. RESULTS: The agreement between measurement and MC is shown to be very good: Within +/- 2% for the PDD and within +/- 3.5% inside the radiation field for all the collimators with the F1 bowtie filter. A full account of the absolute calibration method is given and dose calculation is validated against ion chamber measurements in different locations of a plastic phantom. Calculations and experiments agree within +/- 2% or better in both at the center and the periphery of the phantom, with worst agreement of 4.5% at the surface of the phantom and for one specific combination of collimator and filter. Patient dose from CBCT scan reveals that dose to tissue is between 2 and 2.5 cGy for a pelvis or a lung full acquisition. For H&N dose to tissue is 5 cGy, with the unit presets used in this work. Dose to bony structures can be two to three times higher than dose to tissue. CONCLUSIONS: The XVI CBCT unit has been fully modeled including the F1 bowtie filter. Absolute dose distribution from the unit has been successfully validated. Full MC patient dose calculation has shown that the three-dimensional dose distribution from CBCT is complex. Patient dose from CBCT exposure cannot be completely accounted for by using a numerical factor as an estimate of the dose at the center of the body. Furthermore, additional dose to bone should be taken into account when adopting any IGRT strategy and weighed vs the unquestionable benefits of the technique in order to optimize treatment. Full three-dimensional dose calculation is recommended if patient dose from CBCT is to be integrated in any adaptive planning strategy.

High-performance computing for Monte Carlo radiotherapy calculations.

We report on the RTGrid project, which investigates approaches for using high-performance computing infrastructures, such as the grid, in order to reduce the turnaround time of Monte Carlo (MC) simulation-based radiotherapy treatment planning. The main aim of this project is to render accurate dose calculations using MC simulations clinically feasible. To this end, we have successfully implemented and deployed the RTGrid distributed simulation framework for MC dose calculations. In this paper, we present the main experimental findings.

Simulating oblique incident irradiation using the BEAMnrc Monte Carlo code.

A new source for the simulation of oblique incident irradiation has been developed for the BEAMnrc Monte Carlo code. In this work, we describe a method for the simulation of any component that is rotated at some angle relative to the central axis of the modelled radiation unit. The performance of the new BEAMnrc source was validated against experimental measurements. The comparison with ion chamber data showed very good agreement between experiments and calculation for a number of oblique irradiation angles ranging from 0 degrees to 30 degrees . The routine was also cross-validated, in geometrically equivalent conditions, against a different radiation source available in the DOSXYZnrc code. The test showed excellent consistency between the two routines. The new radiation source can be particularly useful for the Monte Carlo simulation of radiation units in which the radiation beam is tilted with respect to the unit's central axis. To highlight this, a modern cone-beam CT unit is modelled using this new source and validated against measurement.

Characterization of a 2D ion chamber array for the verification of radiotherapy treatments.

In this study we investigated the characteristics of a commercial ion chamber array and its performance in the verification of radiotherapy plans. The device was the 2D Array Seven29 model (PTW, Freiburg, Germany). This is a two-dimensional detector array with 729 ionization chambers uniformly arranged in a 27 x 27 matrix with an active area of 27 x 27 cm(2). The detector short-, medium- and long-term reproducibility have been tested through an extensive set of repeated measurements. Short-term reproducibility was well within 0.2%. Medium- and long-term reproducibility were within 1%, including set-up reproducibility errors and linac output fluctuations. Dose linearity was also assessed. The system response to dose was verified to be linear within the range 2-500 MU. Output factors matched very well pinpoint chamber measurements performed in the same experimental conditions with a maximum local percentage difference of 0.4%. Furthermore, the 2D Array sensitivity to millimetric collimator positional changes and to perturbation effect of irradiated area was tested. The comparison with ion chamber data carried out in water was very satisfying. Finally, measurements of wedge-modulated fields and IMRT beam sequence matched very well ion chamber dose profiles acquired in a water tank. The extensive tests performed in this investigation show that the 2D Array Seven29 is a reliable and accurate dosimeter and that it could be a useful tool for the quality assurance and the verification of radiotherapy plans.

Correction for dose-response variations in a scanning liquid ion chamber EPID as a function of linac gantry angle.

The response of electronic portal imaging devices (EPIDs) of the scanning liquid ionization chamber (SLIC) type is known to vary with linear accelerator gantry angle. This work considered several contributing factors, quantified the artefacts, monitored their reproducibility and investigated the effects of repeated gantry rotations. Unflatness of up to 5% was found. A correction technique was devised using nonlinear regression of a three-variable sinusoidal modulation. Comparison with two existing techniques found our method to be the most effective, providing a flatness well within 2%. This improved accuracy is expected to benefit more accurate dosimetric studies in particular. The post-acquisition correction process required no change in imaging protocols. Applicability of the new technique was demonstrated on images acquired on different days and with different beam sizes. Since the artefacts compromise both accurate dosimetry and image quality, their successful removal should benefit a broad range of SLIC EPID applications.